Speaker:

• • Christopher Upkes, Neo4j
• • Simon Floyd, Industry Director, Manufacturing, Google Cloud

Session type: Full Length

Abstract: This presentation will outline the GCP-approved, event-driven architecture and example graph design patterns designed for and used by manufacturing across numerous verticals. You'll learn about best practices defined by both GCP and Neo4j for designing, deploying, and managing enterprise manufacturing digital twin implementations. You'll also get details related to the design and deployment of GCP event-driven discrete manufacturing digital twin analysis platforms that include graph analytics for discrete manufacturing hosted by Neo4j. Also covered in the pressentation are details including the deployment and use of GCP architecture and the specific review of specific design patterns related to discrete manufacturing. By the end of this presentation, you'll have a basic understanding of how to apply the technology and expertise of GCP, including Neo4j, to drastically improve the analysis currently available to the manufacturing world.

Speakers:

• • Dr. Maya Natarajan, Senior Director, Product Marketing, Neo4j
• • Dr. Jesús Barrasa, Senior Director, Sales Engineering EMEA, Neo4j

Session type: Full Length

Abstract: Knowledge graphs are driving industry disruption and business transformation by bringing together previously disparate data, using connections for superior decision support, and adding context to AI applications. In this session, we'll walk you through the fundamental elements of knowledge graphs based on our recent customer experiences and successes. The majority of Neo4j knowledge graph use cases fall on a spectrum of the three major categories of data management, data discovery, and data analytics. Neo4j characterizes these as Data Assurance Knowledge Graphs for data management, Insight Knowledge Graphs for data discovery, and Decisioning Knowledge Graphs for data analytics. Each of these use cases will be discussed in detail using customer success stories – Data Assurance Knowledge Graphs at UBS for increased trust and explainability, Insight Knowledge Graphs at AirBnB for complete visibility and improved productivity, and Decisioning Knowledge Graphs at Boston Scientific for better predictions and more breakthroughs – to showcase knowledge graphs for contextual AI. Attend this session to learn how leading companies are using knowledge graphs and walk away with practical insights on how to build knowledge graphs.

Speaker:

• • Dr. Alicia Frame, Senior Director of Product Management, Data Science, Neo4j

Session type: Full Length

Abstract: In this session, we'll cover the features of graph data science – what it is, how it solves your most daunting business problems, and how Neo4j helps. You'll learn how Neo4j's leading and robust Graph Data Science platform offering solves the pain of scaling and productionizing your workflows, while also making it super simple to get started. Additionally, to give you a more complete picture of the possibilities of graph data science, we'll walk through a few customer success stories that highlight how it has made a difference.

Speaker: Martin Junghanns, Senior Software Engineer, Neo4j

Session type: Full Length

Abstract: Dive into the internals of Neo4j Graph Data Science and see how to achieve the scale you need. You'll learn the implementation choices available for Neo4j Graph Data Science that enable customers to run graph algorithms on large-scale graphs containing billions of nodes and relationships. Additionally, we'll take a look at our compressed graph representations and a selected set of algorithms implementations that make it possible. The talk will also contain results of our internal scalability testing, which will show how GDS allows customers to import large-scale data and get algorithm results within minutes.

Speakers:

• • Zach Blumenfeld, Data Science Product Specialist, Neo4j

Session type: Full Length

Abstract: One of the most important aspects of a data science tool is how easy it is to integrate with other tools. For Neo4j Graph Data Science, ecosystem integrations is a core product principle, and we always strive to "meet data scientists where they are." Join us to learn about our data science connectors, integrations, and partnerships, including best practices and reference architecture. Start powering your insights with graph data science today!

Speakers:

• • Anurag Tandon, Senior Director, Product Management, Neo4j
• • Junxiang Chen, Engineering, Neo4j

Session type: Full Length

Abstract: In this talk, we take a look at how you can use Neo4j’s developer tools – Data Importer and Browser – to get you up and running quickly on your next project.

Speaker: Luke Gannon, Product Manager, Neo4j

Session type: Full Length

Abstract: Want to run graph analytics and machine learning with zero administrative overhead? Neo4j AuraDS is a fully managed SaaS offering, providing the easiest way for data scientists to get started in the cloud.

Speaker: Mark Baker, Product Manager, Neo4j

Session type: Full Length

Abstract: There are over 200 different AMIs available in AWS from sources such as Community AMIs and AWS Marketplace that purport to provide a Neo4j graph database. In addition to Neo4j Enterprise and Neo4j Community editions, there are container images and Helm charts, as well as AuraDB Enterprise, Neo4j's fully managed graph database platform. Add the 100+ AWS services that you might want to integrate with Neo4j , the almost infinite AWS configuration options for storage, networking, instance selection, and security, and a developer easily becomes overloaded with choice. In this session, we'll look at the different ways to get started with Neo4j in AWS so you can get the best setup to meet your personal, business, and budgetary needs. We will also look at getting data into Neo4j on AWS so you quickly get productive with your graph model.

Speaker: Nathan Smith, Senior Data Scientist, Neo4j

Session type: Full Length

Abstract: Congratulations, your Graph Data Science Community Detection Algorithm has completed successfully! Now, how should you interpret the results and explain them to your colleagues? In this session, you'll learn how statistics such as modularity, conductance, and clustering coefficient can help you decide if your communities are cohesive enough to be meaningful. They can also help you choose the most meaningful result from the output of multiple differently configured Community Detection Algorithm runs. We will also look at ways to describe the communities that emerge from Community Detection Algorithms, which includes looking at the distributions of property values and finding the nodes that are most central within each community.

Speaker: Mark Quinsland, Senior Field Engineer, Neo4j

Session type: Full Length

Abstract: Software-defined vehicles are a subset of Digital Twins that are used by many auto OEMs, Tier-1 Suppliers, and other manufacturers to help deal with supply chain disruption, product management, warranty issues, customer-360, and validation / testing. We will explore several real-world use cases that take advantage of Neo4j's ability to join many disparate types of data to add the context needed for decision makers.

Speakers:

• • Dr. Dave Rench McCauley, Senior Director, Data Science and Machine Learning, Ernst & Young - Quantitative Scientific Solutions
• • Dr. Michael Smith, Lead Scientist, Quantitative Scientific Solutions

Session type: Full Length

Abstract: In technical consulting, we help clients by building them cutting-edge tools to solve seemingly intractable problems. However, setting up infrastructure and building models is only part of the solution when it comes to providing value to customers. Visualization is the last-mile element that can make a client happy or leave them dissatisfied. In this talk, we will cover how we used Neo4j, Graph Data Science, advanced natural language processing, and the new Neodash visualization tool to track flows of scientific knowledge over time for our client, the U.S. National Science Foundation. We’ll discuss how we used language embeddings and automated clustering optimization to model the language of scientific articles (the “content”) and their positions within a citation network (the “context”) to model the evolution of ideas over time, at a scale that would be impossible with other tools. We will also cover how our collaboration with Neo4j, and specifically with the developer of the Neodash tool, enabled us to add the final ingredient of a successful project for this client: a front-end that allowed us to tell them an interesting and useful story while also providing them with a tool to realize the utility for their own work.

Speaker: Spencer Shiotani, Principal Cognitive Software Engineer, Northrop Grumman

Session type: Full Length

Abstract: Maximizing efficiency in facility planning is difficult because buildings are complex systems. Tracking components such as electrical systems, security, network and data, people, seats, doors, etc., can become an insurmountable task. We propose a novel strategy to grapple with this inherent complexity through the utilization of graph-based Digital Twins in Neo4j. Producing a Digital Twin involves many different data sources, including architectural diagrams, HR data, and LiDAR scans to capture 3D representations of building interiors. In this talk, we'll demonstrate how Northrop Grumman creates and implements these Digital Twins to orient new team members, optimize seating assignments, communicate changes to the employee base, track resource usage, coordinate large-scale moves, and analyze the delta between what was planned and what actually exists, thus making a historically underutilized resource more accessible and manageable.

Speaker: Alexander Koderman, Senior Developer, SerNet, Inc.

Session type: Full Length

Abstract: State-sponsored and state-tolerated cyber attacks continue to rise. Governments and regulators also continue to respond. Companies are facing an increasing number of compliance requirements and controls. The result is that assessment cycles are becoming faster and control satisfaction needs to be verifiable with high granularity down to single control statements for individual systems or even system components. The U.S. National Institute of Standards and Technology (NIST) has developed OSCAL, a machine-readable language for cybersecurity control implementation and assessment. The next step is to develop implementations to aid cybersecurity practitioners in their daily tasks, such as: determining control prerequisites, finding related controls, tailoring controls to the organization,and assessing control implementation. We demonstrate "OSCL4NEO4J" – a set of open source scripts and REST API that can be used to import and work with OSCAL data in Neo4j to solve practical problems faced by cybersecurity practitioners every day. The open source project that we present has already been recognized by NIST.gov and is referenced from their official OSCAL repository.

Speaker: M. David Allen, Senior Director of Developer Relations, Neo4j

Session type: Full Length

Abstract: Neo4j & Neo4j AuraDB support Python, JavaScript, Java, Go, and .NET. In this session, we’ll cover some best practices for using Neo4j drivers in your application. We'll provide worked code examples of the most common things people try to do, and good patterns that will making coding with Neo4j easier. We will also dive into how querying a Neo4j database works, how Neo4j clusters operate, how queries are sent through the system, and what's happening under the hood of how drivers work.

Speaker: Thomas Silkjaer, Head of Analytics and Compliance, XRP Ledger Foundation

Session type: Full Length

Abstract: A story of ups, downs, learnings, and findings from working with representing the XRP Ledger in Neo4j. What started as a hobby project in 2018 to represent and analyze payments only, expanded in 2019 into a full history graph representation of the XRP Ledger blockchain that has been running for more than nine years with 2.2 billion transactions, generating more than 1.5 billion ledger objects. In 2022, the data model is updated to better scale with increased XRP Ledger use, reduce the storage footprint more than 50 percent by applying learning from the past years to remove redundant properties, move unused data to JSON strings that can be parsed with APOC as needed, and to reflect new possibilities with Neo4j 4.x. This talk also showcases how the database, that is kept in sync with the XRP Ledger +/- 10 seconds, is used in the fight against criminal finances by “following the money,” and how it is used to stay ahead of money laundering when criminals move funds quickly around prior to moving it to legitimate exchanges.

Speaker: Dr. Janez Ales, Senior Research Scientist, Mathematician, Data & Algorithms, Knowledge Architecture & Innovation, BASF

Session type: Full Length

Abstract: World's Journal and Patent Knowledge Graph on 55 billion entities offers endless opportunities for traversals at scale, with challenges in data size and traversal execution times. However, todays limits of available cloud hardware speeds can be reached with algorithms on much smaller data sets due to combinatorial complexity. We discuss combinatorial tasks and challenges hidden in many industry data sets and decision problems, and highlight differences and benefits of graph database direct access via Cypher tools vs. language APIs on various use cases.

Speakers:

• • Danai Eleni Aristeridou, Data Scientist, Pfizer, Inc.
• • Anastasia Karatzia, Data Scientist, Pfizer, Inc.

Session type: Full Length

Abstract: Harnessing healthcare data and unraveling the steps until the diagnosis of a chronic disease can be a challenging task. Revealing the path to a chronic condition could potentially lead to the early diagnosis, support treatment adherence, and adverse events mitigation. Every patient follows a unique path until the diagnosis of the target disease. Electronic health records and claims contain information about diagnoses, prescriptions, procedures, hospital admissions, and so on. Graph theory could capture relationships between the various data records, facilitate the identification of common paths between patients, and point out key comorbidities. In this session, we present a framework for the mapping of the patient journey by applying graph theory. The target is to better understand the progression of a chronic disease and reveal concealed connections between diseases that may not be visible with traditional visualization and predictive modeling techniques. Finally, we present how this framework can be transformed into a tool to quickly provide insights and recommendations to the decision makers.

Speaker: Jennifer Reif, Developer Advocate, Neo4j

Session type: Full Length

Abstract: Many of us have built applications for traditional data structures (like relational database tables), but is it different for graph data stores? Do developers need to retool and relearn? In this session, we'll cover a brief introduction to graphs, walk you through writing a typical Java application with Spring, and connect it to Neo4j. From interacting with the graph data from the application to deploying to the cloud, you'll see the process from start to finish. You'll also learn how to tackle pitfalls and pick up tips along the way, as well as explore the ways we can build, deploy, and connect applications to the database. This will come alive through a live demo, as we see the results of our efforts. Come to this session to build your business applications for graph data!

Speakers:

• • Daniel Kang, Senior DevOPs Engineer, Adobe
• • Gabriel Tucker, SR Data Architect, Adobe
• • Daniel Vilajeti, DevOps Engineer, Adobe

Session type: Full Length

Abstract: Did you ever think you could setup your Casual Cluster to be self-healing and auto recoverable in the cloud? Would you like to know that your backups will restore without error and that your data is consistent every day? Come to this talk to learn more about running a stress-free Neo4j Causal Cluster. We'll also cover automated backups, daily restore testing / data consistency check, autoscaling groups, Flatcar OS, Docker implementation, Ansible, Terraform, ELB endpoints for leader, followers, and read replicas, ENI for persistent IP, as well as how to select the right instance types, gotchas, and the benefits of upgrading to Neo4j 4.4.

Speakers:

• • Sal Aguinaga, Master Data Scientist, Deloitte | AI Center of Excellence
• • Dr. Sanmitra Bhattacharya, AVP, Data Science, Deloitte | AI Center of Excellence

Session type: Full Length

Abstract: Each year billions of insurance claims are submitted by healthcare providers. U.S. healthcare spending continues to grow over five percent year-over-year and accounts for approximately 20 percent of the Gross Domestic Product. The National Health Care Anti-Fraud Association conservatively estimates healthcare fraud at three percent of total health care costs, which in 2019 represented over a hundred billion dollars in fraud. The Centers for Medicare & Medicaid Services and other regulators mandate fraud, waste, and abuse (FWA) surveillance by payors of healthcare claims. Screening providers based on their risk profiles across various dimensions of FWA is a key component of such surveillance. Our project identifies providers sharing common risk signatures with other providers – uncovering pairwise similarity using graph-theoretic algorithms and graph neural network (GNN) methods. This two-pronged solution works with Neo4j’s graph engine at its core by applying Graph Data Science and serving quality graph datasets to external state-of-the-science GNN training workflows. The objective of these two approaches is to produce complementary groupings of providers with common risk signatures. Our analyses reveal the likelihood of hidden or unknown relationships between providers across various FWA dimensions.

Speaker: Gal Engelberg, Research Associate Principal, Accenture

Session type: Full Length

Abstract: Today, enterprises in general and industrial manufacturers in particular are increasingly connecting to external networks. As such, industrial processes that were once isolated from the open internet network are now more vulnerable to external cyber attacks. As the frequency and resulting impact of these vulnerabilities increases, there is a need to prioritize and mitigate risks in order of importance to the business. Unlike common risk assessment tools that prioritize risks based on their potential damage to the infrastructure layer alone, we add the business context to the equation. Using Neo4j, we present a knowledge-graph-driven approach to address the above challenges. Our work will be demonstrated over a vehicle assembly smart manufacturing environment. First, we present the notion of process-aware attack-graphs: a semantic representation of the factory infrastructure and industrial-process layers. We base the approach on the usage of graph data science algorithms to quantify the cybersecurity risk based on potential adversary behaviors. Then, map the risk from the infrastructure layer to the process layer. And lastly, to identify the risk root cause and recommend, which issues to address first accordingly. This session will be focused on the usage of Neo4j Graph Data Science algorithms over knowledge graphs while triaging business and cybersecurity.

Speaker:

• • Andres Ortiz, Engineering, Neo4j
• • Darrell Warde, Engineering, Neo4j

Session type: Full Length

Abstract: Join Darrell and Andrés from the GraphQL Team at Neo4j as they talk about one of the newest features of the Neo4j GraphQL Library: GraphQL Subscriptions. Using this new feature, GraphQL API consumers can listen to data changes in real time, which happen in Neo4j via the GraphQL Library. Following a high-level overview of the Neo4j GraphQL Library, they will demonstrate the new Subscriptions feature. You can also expect a deep dive of how it works under the hood.

Speaker: Michael Hunger, Senior Director, User Innovation, Neo4j

Session type: Full Length

Abstract: Cypher is a great, powerful query language, enabling you to quickly get the most from your graph queries. For some, squeezing out every bit of performance is necessary, and you wish you had the capability to do so packed into a reusable LEGO block for your queries. In this session, you'll learn from practical examples of how to build user-defined procedures, functions, and aggregation functions with just a few lines of Java code. We'll dive into how to efficiently use the Java API and what to watch out for to ensure you get the most out of your work. This will help you to build your own extensions to Neo4j or understand better how the existing procedures and functions work under the hood.

Speaker: Darrell Warde, Engineering, Neo4j

Session type: Lightning Talk

Abstract: With Neo4j as the graph database, the GraphQL Library makes it simple for applications to have application data treated as a graph natively from the front-end all the way to storage, avoiding duplicate schema work and ensuring flawless integration between front-end and backend developers. In this session, we'll walk through how to use GraphQL with your existing Neo4j database.

Speaker: Eric Monk, Principal Solutions Engineer, Neo4j

Session type: Lightning Talk

Abstract: New to Fabric? Join Eric Monk for a quick introduction on how to get started with Fabric. The session will briefly breakdown important Fabric concepts and then show you the steps of how to configure the Fabric database, setup users, and perform a Cypher query that queries an Aura database and a local database using Fabric.

Speaker: Cynthia Femano, Senior Solutions Architect, Neo4j

Session type: Lightning Talk

Abstract: Graph databases and graph data science tools bring the connections and context needed to make AI / ML algorithms work better, revealing insights embedded deep within your data. In today's world of highly connected data, the use of graph algorithms to supplement traditional data science methods can help you uncover knowledge that may be overlooked otherwise. This presentation will introduce you to the topic of Neo4j Graph Data Science at a high level, why it deserves your attention, and will leave you wanting to learn more. This is a wave you most definitely want to jump on... don't be left behind!

Speaker: Yi Ren Sum, Software Engineer, Neo4j

Session type: Lightning Talk

Abstract: Data represented as graphs usually contain some wonderful stories. Using graph data science, one can calculate interesting insights about this data. Without the final step of conveying these insights, however, the true color of data is often under-appreciated. In this lightning talk, we'll show you some ways to visualize data using Neo4j Bloom and Neo4j Graph Data Science.

Speakers:

• • Arthur Zverko, Software Engineering Team Lead, EquityBee
• • Gal Bello, Senior Field Engineer, Neo4j

Session type: Lightning Talk

Abstract: EquityBee is a growing startup, and as such, scaling up and expansion is a super crucial step for their business success. The unique use case has placed a harsh demand for the ability to expand and scale. The decision was to transit to a graph database and present a custom ORM within it. The ORM was originally built to define all schemas, relationships, and data validations. This transition was made with harsh performance requirements, and that is why EquityBee aimed to keep the NodeJS codebase clean and performant – while making the dev team’s life easier – by using a custom-built Query Builder to enable developers to code in a JS object manner. Presented by EquityBee's Founder and CTO, this session will focus on their unique use case and how they have scaled their operational database from Firebase to Neo4j AuraDB.

Speaker: Krishnan Narayan, Distingushed Engineer, Palo Alto Networks

Session type: Lightning Talk

Abstract: Prisma Cloud from Palo Alto Networks is a leader in cloud security, securing over 1B+ assets and providing the most comprehensive enterprise security solution for cloud users. One of the unique challenges with security posture detection in a deeply connected ecosystem like the public clouds of AWS, Azure, GCP, etc. is to be able to actually use these relationships to uncover advanced threats to the infrastructure that may otherwise go completely unnoticed. Join Krishnan Narayan for a quick overview of how PaloAlto Networks apply graph theory to solve some of the most advanced challenges around security posture detection and response.

Speaker: Dr. Edgar Osuna, Chief Data & Analytica Officer, Todo1

Session type: Lightning Talk

Abstract: Digital transformation in the financial industry has been profound in the last decade and continues to accelerate. Unfortunately, fraudsters have also modernized and are exploiting the advantages of the digital landscape to make a profit. This presentation covers the success story of a graph-based machine learning application built to fight digital fraud in real time that is currently helping a wide array of financial institutions.

Speakers:

• • Alexander Erdl, Senior Director Marketing Manager, Neo4j
• • Michael Hunger, Senior Director, User Innovation , Neo4j

Session type: Lightning Talk

Abstract: Ever wondered about the background of the audience around you at a conference? At GraphConnect 2022, we will show you! At registration every participant is sharing some details about themselves and during our session we will showcase where people are from, what favorite programming languages they use, and more. We'll also import the agenda of the conference into Neo4j and make it accessible in Neo4j Bloom. In this session, we want to showcase how easy it is to go from model to graph. We'll start with a concept, with the Neo4j Data Importer, and then load data into Neo4j and explore this data in Neo4j AuraDB Free and Neo4j Bloom. By including data from the audience, we make the results more tangible and interesting for everybody. We can host the graph afterwards on AuraDB for everybody to explore (at least the schedule part).

Speakers:

• • Erik Bijl, Data Scientist, Rabobank
• • Salomon Tetelepta, Data Scientist, Rabobank

Session type: Lightning Talk

Abstract: As Rabobank, we are striving to apply existing risk assessment with insights from network analytics. As a first step, we launched a project aiming at detection of clients participating in money laundering schemes. This project will be the main topic of our presentation. In this project, the central question was how to determine whether a client is actively participating in money laundering by using network analytics. The network was captured by extracting network features. These features were a result of running a variety of graph queries and Algorithms on top of our graph model. In this presentation, we will discuss how we build up the graph data model, how we apply graph queries, and which algorithms can be run on these graphs. Sharing our lessons learned, we hope to give you valuable insights on this topic.

Speakers:

• • Jeff Gagnon, Product Manager, Neo4j
• • Sebastian Wictorin, Engineering, Neo4j

Session type: Lightning Talk

Abstract: Deriving insights from data is a naturally inquisitive process grounded in exploration. Neo4j Bloom is graph data visualization software built for investigation, exploration, and collaboration without requiring any coding experience. Join Jeff and Sebastian as they show how Bloom’s easy-to-use, illustrative design and powerful graph analytics help you to paint a beautiful picture of what your data is telling you.

Speakers:

• • Florent Biville, Developer, Neo4j
• • Eric Solender, CTO, MindStand
• • Nikita Wootten, Chief Data Scientist, MindStand

Session type: Workshop

Abstract: In this session, learn the different ways to interact with Neo4j from Go. We will go over usage of the official Go driver, including connecting to the database, running queries, and managing transactions. We'll then explain how using an Object Graph Mapper (aka OGM) like GoGM can simplify the process of interacting with Neo4j.

Speaker: William Lyon, Developer Advocate, Neo4j

Session type: Workshop

Abstract: Bring the power of graphs to iOS mobile app development in this hands-on workshop. We will explore how to use the Neo4j GraphQL Library to build GraphQL APIs backed by Neo4j AuraDB and how to integrate GraphQL into an iOS app using SwiftUI. Some familiarity with Swift and iOS app development will be helpful, although not required. To follow along during the workshop attendees will need a Mac laptop with a recent version of Xcode installed.

Speaker: Dr. Jesús Barrasa, Senior Director Sales Engineering EMEA, Neo4j

Session type: Workshop

Abstract: This workshop is a hands-on lab with an ambitious goal: show three practical examples of using ontologies in Neo4j, step by step, in a way that attendees can follow and understand or even join and code along (on their local Neo4j or in a Sandbox). The three examples are as follows. Data import: Use an ontology as the target model for the graph that your ETL pipeline will build in Neo4j. Model validation: Use an ontology to define constraints on the shape of your graph and create a data quality dashboard reporting on the violations of those constraints in your Neo4j graph. Semantic search: Use the ontology to semantically annotate a dataset and implement semantic search and semantic similarity using Cypher over your data and the ontology. For this workshop, we will use public ontologies in different domains (FIBO, Schema.org, ESCO...) and open datasets that will be shared day of with attendees.

Speakers:

• • Alistair Jones, Director of Engineering, Neo4j
• • Irfan Nuri Karaca, Staff Software Engineer, Neo4j

Session type: Workshop

Abstract: Learn how to draw graphs using the arrows.app tool from Neo4j Labs. In this hands-on session, you'll learn the quickest way to create a simple graph in your web browser and then use this graph to develop a data model for your graph application. We'll run through a series of interactive exercises that will teach you how to draw simple graphs, import text, import from a spreadsheet style and format, share graphs, export images, develop a data model, use powerful editing features, refactor graphs, import into a Neo4j graph database, and run algorithms on imported data.

Speaker:

• • Thomas Larsen, Forensic Analytics Manager, ABB Inc

Session type: Lightning Talk

Abstract: Discover how a newly-formed Forensic Analytics team at ABB got started on their graph journey. This session will share how the team moved from a vague idea to a critical tool used daily by its members. You’ll learn how the team was able to seamlessly integrate the tool into their existing data pipeline without additional, burdensome overhead.

Speakers:

• • Gaëtan Belbéoc'h, Head of Product - Digital Experience Platform, Adeo
• • Charles Gouwy, Product Owner Knowledge Graph for Publication, Adeo

Session type: Full Length Session

Abstract: This session will explore how Adeo improves the experience of its e-commerce sites thanks to the knowledge graph. You’ll learn about the design of the semantic digital platform developed by Adeo, what its main components are, especially the Neo4j knowledge graph, and how they contribute to improving the customer experience of the brand's e-commerce sites.

Speakers:

• • Ido Faran, Software Team Leader, AppsFlyer
• • Olga Kogan, Software Architect, AppsFlyer

Session type: Workshop

Abstract: Join us to learn how we built an innovative centralized policy-based authorization solution for microservices architecture using Neo4j and OPA (open policy agent). We will explore the advantages and challenges of having a shared repository for entities and an authorization data model. Our business-critical domain model, which provides the context to all AppsFlyer data analytics products, is managed in Neo4j and reflects the entire graph of relationships between our customers and partners. In addition, we use Neo4j as the control plane repository of the authorization service. Managing permissions configuration and entities as part of the same graph allows us to easily manage the lifecycle dependencies between the two. Our innovative authorization solution, on top of Neo4j, allows us to easily change the authorization logic when new functionality is added or when data sharing policies or privacy regulations change. We help more than 80K customers by providing advanced analytics solutions and products for marketing teams and application owners. Our customers include global brands such as Coca-Cola, Nike, Pinterest, Visa and more. We also integrate with over 9K partners, including media partners, customer engagement platforms, campaign management platforms and more.

Speakers:

• • Mike Morley, Director AI/ML Technology, Arcurve
• • Peter Tunkis, Lead Data Scientist, Arcurve

Session type: Workshop

Abstract: In this presentation, we will demonstrate how graph database technology and 3D modeling software can be utilized to address a wide variety of business questions and challenges. We take a basic example of modeling a physical system (engineering design model for building ventilation) using Neo4j and Graph Data Science to project a visualization of the physical model into a AR/VR-style environment and help users track and predict contaminate transport. This approach is designed to be versatile and can be applied across contexts.

Speaker: Saif Ur-Rehman, Senior Data Engineer, Basecamp Research

Session type: Lightning Talk

Abstract: At Basecamp Research, we believe that biodiversity is our greatest asset, and that it should be valued as such. Through biotechnology, biodiversity offers the first credible chance to wean our world from petrochemistry towards a cleaner, biochemistry-based future. We are building a bridge between biodiversity and the bioeconomy – transforming the connection between these communities. By doing this, we are creating a new, nature-based value chain that simultaneously promotes protection of the Earth’s wild places and provides the building blocks for a cleaner, healthier, and more sustainable future for all. We use Neo4j to draw links between the genomic and taxonomic contents of samples, collected from around the world, to allow discovery of commercially valuable products for the bioeconomy that are identified through our ML pipeline and graph algorithms. With Neo4j, we help ensure that the benefits from these products flow back to the stakeholder communities and guardians of biodiversity.

Speaker: Julian Schibberges, DIrector, Bernstein Analytics GmbH

Session type: Full Length Session

Abstract: Understanding political interests is key in a democratic society and graphs help us understand those interests better. Based on the newly-introduced lobby register of Germany's federal parliament, we show how structuring data as a graph yields new insights into the networks of political advocacy. Using Neo4j Graph Data Science we dive even deeper, identifying hidden actors and using link prediction to connect lobbyists with parliamentary initiatives. "

Speaker:

• • Cyrine Kaabachi, Senior Data Scientist, BNPP

Session type: Full Length Session

Abstract: Banking information systems process various types of transactions and produce heterogeneous data. Such IT data can be used to extract insights and help IT Operations in their daily activities. It is a major need to build data-driven solutions for information system management and support IT Ops. Large scale industries have critical issues and incidents that need to be resolved in a timely manner. Our goal is to provide data science based solutions and ITOps tools to speed up incident root cause analysis and risk assessment. To achieve this, we combine graph techniques and embeddings on unstructured IT data to make complex correlations. Our final solution is based on a recommendation system using graph features to speed up critical issues investigations. This talk shares a use case based on real-world application of graph data science to address IT and business goals.

Speakers:

• • Matt Butler, Co-Founder, Bonsai
• • Corey Lanum, Chief Product Evangelist, Cambridge Intelligence

Session type: Full Length Session

Abstract: Most marketing analytics tools focus on aggregate views of customer behavior: a webpage had 1000 unique visitors, an email was opened 200 times, a campaign generated 50 new leads. But any marketer will tell you that modern customer behavior is rarely linear. Aggregated analysis can never tell the full story, because the customer journey isn’t a straightforward model – it’s a graph. At Bonsai, we built a marketing analytics platform based on that principle. Using Neo4j, it maps out a business’s touch points and plots the varied paths between them. The Bonsai platform combines Neo4j Bloom and KeyLines by Cambridge Intelligence. Marketers can visualize typical customer journeys and analyze each interaction, its timing and sequence, to understand which links in the chain add incremental value – and which need improving. In this session, we will share our experience at Bonsai of building the platform on Neo4j Bloom and KeyLines, and explain how modeling our data as a graph has transformed the way marketers think about their customer journey. Corey Lanum of Cambridge Intelligence will also demonstrate how graph visualization enhances the platform by making complex customer data accessible and understandable to business users.

Speaker:

• • Sébastien Tourlet, Director, Data Science & Engineering, Cap Gemini

Session type: Full Length Session

Abstract: KG4P uses an integrated technological stack including Neo4j, Dataiku, and Linkurious to build, refactor, exploit, and visualize a multi-source, life science knowledge graph. This graph is used to identify biomarkers for multiple myeloma. The Neo4j, Dataiku, and Linkurious Partnership as a product offers an all-in-one technical stack to accelerate data-driven use cases and projects. This versatile and agile solution eases the highlighting of hidden results and opportunities (new drug targets, strong biomarkers, new indicators, best drug responders, new collaborations, etc.). KG4P leverages data and insights and breaks data silos using graph technology – an all-in-one technical stack to cross data and generate operational insights.

Speaker: Leon Goldberg, Chief Architect, Cider Security

Session type: Full Length Session

Abstract: CI/CD environments and processes are increasingly becoming a key area of focus for hackers and consequently, defenders as well. However, truly visualizing the attack surface has become a complex engineering task as the number of exploits and vulnerabilities grow daily. This talk will walk you through the research performed to create an improved CI/CD graph model for hackers and defenders to provide greater control and defense mechanisms for the most central part of today's engineering organizations. There were many engineering complexities involved when modeling the graph, but ultimately the goal was to provide greater observability at scale, improved attack vector blast radius estimations, more precise defense simulations, and easier discoverability of potential attack vectors. Join us on the journey of modeling the graph to learn how to protect and dramatically improve visibility into the many nodes and edges of your most lucrative systems.

Speaker: Alvaro Oslé, Director of Architecture, Ciena - Blue Planet

Session type: Full Length Session

Abstract: Attend this session to learn about how Neo4j can be used to model telecom networks, with a deeper focus on how to automate the network planning, design, and service assurance functions.

Speakers:

• • Jason Schatz, Principal Software Development Engineer, CodeLogic
• • Rob Vrooman, Principal Software Development Engineer, CodeLogic

Session type: Full Length Session

Abstract: Making informed development decisions requires a strong understanding of the connections and complexity within and across your application landscape. With software changing faster than ever, and dozens of applications to manage within a single enterprise, it is often difficult to have a clear view of how everything fits together. CodeLogic equips engineers with the most comprehensive software dependency data available, combining binary and runtime scanning to create a complete graph of an application’s structure. In this session, you will learn how CodeLogic utilizes Neo4j and CypherQL to capture and analyze data achieved through application profiling, and how their data model visually mimics the source code itself. Jason Schatz (Principal Software Development Engineer, CodeLogic) and Rob Vrooman (Principal Software Development Engineer, CodeLogic) team up to discuss how moving from a relational database to a Neo4j graph database gave them the ability to visualize and distill complex codebases quickly. Attendees will learn a rare use case for graph and see how the CodeLogic backend models data into simplified maps that can be easily analyzed to identify cross-application dependencies, navigate code change impact, and ultimately reveal the bigger picture.

Speaker: Tawanda Ewing, Machine Learning Engineer, Deep Learning Cafe

Session type: Lightning Talk

Abstract: We live in a time where large volumes of data are generated on a daily basis, and one challenge, as a result of these volumes, is effectively exploring the data. Finding methods to do so helps us derive useful insights that lead to us making better decisions within an organization or as a society. Knowledge graphs, which are created in graph databases, are proving to be a method worth noting thanks to their ability to model relationships between our data. One way to appreciate just how powerful they are is by looking at an example we can all relate to – the languages we speak.

Speakers:

• • Andrew Nepogodin, Cloud Architect, Dell
• • Bhanu Naidu, Data Engineer, Dell

Session type: Full Length Session

Abstract: Dell Digital's traditional enterprise architecture couldn't handle data consolidation and pre-assembling needs, which impacted the performance of their Pricing-as-a-Service offering. It needed to manage exponential data growth, plus handle different data types within the pricing domain. Learn how Neo4j graphs helped improve these challenges, plus the processing of their pricing data and approaches used to optimize licensing costs. We will also walk through the migration process from their legacy system into Neo4j and provide an overview of the current production setup and data volume being served.

Speakers:

• • Joseph Hilger, COO, Enterprise Knowledge, LLC
• • Lulit Tesfaye, Partner and Division Director, Data & Information Management, Enterprise Knowledge, LLC

Session type: Full Length Session

Abstract: Data Fabric is one of the hottest solutions in the data world right now. It is seen as the new way to democratize access to data. While the concept makes sense, the real question is how it can work at scale in large organizations. Enterprise Knowledge is implementing a true enterprise-wide data fabric for one of the largest financial institutions in the United States. This is a project against massive datasets that serves an entire organization. Our client has over 350 petabytes of data that provides information to over 10 divisions within the organization. As part of this presentation, we’ll share how our consultants are designing the abstraction layer, implementing governance, and democratizing access to information across the enterprise. We’ll answer questions about how data fabric works, how it scales, and how your organization can implement its own data fabric solution.

Speaker: David Swank, CEO, EnXchange

Session type: Full Length Session

Abstract: Today's Energy industry is fraught with complications, from old data silos, complex "smart" equipment, and massively changing new demand from EV to solar and wind home generation. In this session, hear how graph databases are revolutionizing the way the energy industry connects and powers the world. All types of players in the industry from transmission companies, co-ops, consumers, and equipment providers are part of the "graph" and will become integrated players in the new world for how we all interact with the energy grid. This session will focus on EnXchange's vision to orchestrate and optimize grid operations serving more than 40 million commercial and residential customers across the U.S. From generation to toaster, EnXchange is using Neo4j and Fabric to deliver actionable insights and predictive analytics across hundreds of independent operators.

Speakers:

• • Abhishek Singh, Technical Manager (Digital Health Data Science)
• • Cristiana von Stosch, Assistant Director Data Science Digital Health

Session type: Full Length Session

Abstract: Not all features are created equal. Hence, feature importance determination is one of the most fundamental problems in machine learning. Most feature importance methods rely on the existence of a target feature (response, output, y) to understand the importance of each feature. But if the target feature is not present and we only have the independent features, then only unsupervised methodologies can be applied and feature importance is not easy to calculate. In this session, a novel graph model methodology is proposed to identify the feature importance of datasets without a target feature. In our proposed approach, the target feature columns of seven datasets were hidden, and their independent feature importance was calculated utilizing the proposed approach and another unsupervised method (Gower Distance). After that, the feature importance was calculated using the full dataset (with the target feature) using commonly used supervised methods (random forest and decision tree algorithms). Finally, the rank of the most impactful features was compared for all methodologies. It was concluded that each algorithm delivered a different variable importance rank that may or may not match with the other’s output, but all the approaches have some level of overlap in variable importance order.

Speakers:

• • Nicholas Jones, Software Engineer I, CSX Technology
• • Dean Schaefer, Software Engineer I, CSX Technology

Session type: Lightning Talk

Abstract: CSX Technology is committed to leveraging the power of graph to maintain equipment state. They have a wealth of operational data for rolling equipment movements (locomotives, railcars, intermodal containers/trailers, and End of Train (EOT) devices). CSX Technology is interested in more accurate and timely equipment associations. In this session, they’ll talk about Locomotive to Train Association (LTA), rolling equipment (railcars to trains, rail stations, customers facilities, railroad interchanges), and roadmap (leveraging locomotive GPS to provide granular shipment updates for associated equipment). You’ll learn about their success story with rolling equipment in the CSX Digital Twin and applying the MERGE keyword to create an immutable graph strategy.

Speakers:

• • Raghavendra Sayana, Cloud Automation and Reliability Engineer, Expedia Group
• • Chris Williams, Principal Software Development Engineer, Expedia Group

Session type: Full Length Session

Abstract: "The Expedia Group Platform serves more than 200 travel sites in 70 countries and encompasses nearly 3 million properties. Expedia developed its Entity Key Graph (EKG) using Neo4j, unlimited entity graph traversals such as starting at a known reservation and moving to the associated unit and then to the property and then to the owner. The ability to traverse the business graph in a native graph engine like Neo4j allows Expedia to easily slice off “views” of data that can be used time and time again. This common graph platform allows any view of data to be created with the same underlying graph supporting it, with no extra indexes or complex SQL queries required. In this session, Chris Williams, principal engineer, will outline how Expedia uses Neo4j in conjunction with MongoDB to build flexible and powerful event-driven views of its massive graph. Chris will be joined by Raghu Sayana, staff engineer, who will discuss how Neo4j fits in Expedia’s automated deployment framework to support this and other use cases with minimal hands-on effort."

Speakers:

• • Harish Arora, Managing Director, EY
• • Maxim Ogienko, Senior Manager, Data and Analytics, EY

Session type: Lightning Talk

Abstract: Organizations need to identify ESG risks associated with their business practices and act upon them quickly to preserve their brand value and avoid any non-compliance penalties. Organizations today spend substantial efforts on analysis of reporting standards and frameworks (such as SFDR, TCFD, SASB, CDP, and many others) to make sure that they are not violating any ESG compliance standards or any voluntary ESG commitments. Many of these reporting standards are interrelated and share common data elements; closely inspecting relationships of such disclosures through a purpose-built ESG knowledge graph can assist human disclosure reviewers, reduce total manual review efforts, and significantly speed up the time taken to comply with the ever-evolving ESG reporting landscape. In this session, we will show you how to visualize the ESG (Environmental, Social, and Governance) reporting landscape in graphs.

Speakers:

• • Jill Putnam, Enterprise Data Manager, Federation of State Medical Boards
• • Anne Lam, Sr. Software Engineer, Federation of State Medical Boards
• • Enrique Urrutia, Data Analyst, Federation of State Medical Boards

Session type: Full Length Session

Abstract: Over 30,000 physicians register each year to take the final step of the high-stakes United States Medical Licensing Exam. The Federation of State Medical Boards (FSMB) is the single point of entry for this registration. The complex and arduous journey to becoming a fully licensed physician is incomparable to any other professional requirements, with the health care of the general population dependent on its success. This presentation explores the first use of a Neo4j graph database by the FSMB. It was designed to improve the flow and experience on the exam registration website, guiding physicians through the process. Join us to learn about our journey with Neo4j, including why we decided to use a graph database, our rookie mistakes and deployment triumphs, and the design evolution that led to a custom data ingest tool that we fondly call “Rocket.”

Speakers:

• • Matthew Wallace, Data Architect Lead, Flint Hills Resources
• • Zach Fenton, Manager Enterprise Data and Solutions, Flint Hills Resources

Session type: Full Length Session

Abstract: Solutions Workbench has been an integral part of our organization's implementation and adoption of Neo4j. It allowed us to easily model data and see how well different versions work. Part of the magic that we have seen is that when creating these models, we are able to talk to IT and the business with the same diagram. In previous tools we needed to create different sets of diagrams depending on the audience we were speaking with. Not only can we talk with various levels of our organization with the diagrams created, but we are also able to quickly adjust on the fly to adapt to new business requirements. The tool also pushed us to think about the value that could be created not just with the data we were putting into Neo4j, but the metadata we curate along the way. Which lead to the analysis of our models that helped us paint the picture of just how connected our data is across our various use cases.

Speakers:

• • Scott Ogden, Head of Commercial Data Science, Genmab
• • Sam Wagner, Associate Director, Commercial Data Science, Genmab

Session type: Lightning Talk

Abstract: In this session, we'll share our process to becoming a data-driven organization, focused on deeply understanding our patient and physician journey in the context of medical decision-making. Our customer-360 approach unifies seemingly separate real world, marketing, CRM, and social datasets to create the foundation for our analytics strategy. We can understand drivers of patient activity and create the bedrock of our omnichannel recommendation engine by analyzing the connections created in our knowledge graph. Naturally, we obtain our key opinion leaders and influencers in novel ways so we can create a truly unique Genmab interaction with customers.

Speaker: Alexander Jarasch, Head of Data and Knowledge Management, German Center for Diabetes Research and HealthECCO

Session type: Full Length Session

Abstract: HealthECCO is building a unique solution to combine, annotate, and organize the world's health knowledge and get it into the hands of the right people at the right time. HealthECCO, winner of the the Neo4j Graphie Award in 2021, is a non-profit organization committed to open source software and open access to knowledge in order to improve the quality of guideline usage and to foster innovative global research. The beating heart of their platform is a Neo4j knowledge graph that integrates a growing number of different but related data sets. Their data loading pipelines process each of the data sets, indexing nodes and creating connections to other data sets, as well as annotating text in the data using natural language processing (NLP). The connections in the graph make data findable. These pipelines are portable and repeatable making their data reusable. By incorporating ontologies, they make data interoperable so that data can be reused and repurposed. Our objective is public access for all data sources. Following the FAIR principles (Findable, Accessible, Interoperable, Reuseable), their knowledge graph will not only reveal hidden connections but will be publicly available and globally accessible. In this session, you learn about their diabetes and COVID-19 use cases including Neo4j Graph Data Science applications and several interactive graph applications.

Speaker: Sachin Agarwal, Senior Product Manager, Google Cloud

Session type: Full Length Session

Abstract: Google Cloud Dataflow is a fully managed streaming analytics service for executing Apache Beam pipelines within the Google Cloud Platform ecosystem. Learn how Neo4j's Dataflow template and Apache Beam connector maximize your Google Cloud Dataflow use.

Speaker: Sean Robinson, Lead Data Scientist, Graphable

Session type: Lightning Talk

Abstract: One of the toughest challenges for data scientists adopting Neo4j Graph Data Science is unfamiliarity with Cypher and the Neo4j interface. In this demonstration, we will break down this barrier by demonstrating how to integrate Graph Data Science with Python analytics in Jupyter. Using the GDS 2.0 Python driver, we will work through a graph machine learning use case via Python in Jupyter. We will then integrate and interpret the results using other Python libraries to demonstrate how the Neo4j Python driver offers seamless integration with the tools and libraries data scientists use in their daily work. You will get access to sample code for performing ML and graph analytics in GDS using nothing but Jupyter, Python, and simple Cypher. You will also learn how to integrate your results with popular data science libraries.

Speaker: Luanne Misquitta, VP of Engineering, GraphAware

Session type: Full Length Session

Abstract: Graphs are commonplace in investigative, intelligence, and law enforcement work. One of the primary advantages of a graph is to connect data from various data sources, digital and human, and maximize insights across deep and complex networks of connections, bringing them together in fusion centers for a centralized view of suspicious activities. For analysts, data quality and trust is key. The reliability, validity, and general consistency of data sources that contribute to forming real world fused entities is a factor that influences the analysts’ interpretation of events. This session talks about the challenges related to surfacing these aspects of data provenance and various approaches that can be employed to address them using Neo4j. We will touch on graph modeling, implications for data security, and how sources and information ratings can be effectively shared with analysts who need access to them.

Speakers:

• • Erik Erickson, Chief Data Officer, Hennepin County
• • Alexander Long, Data Engineer, Hennepin County

Session type: Full Length Session

Abstract: As local governments faced the challenges of the COVID-19 pandemic, they were confronted with a significant amount of information. Estimates about the state of the economy, housing, and public safety within communities were readily available, but finding broader insights was much harder. For example, how can we meaningfully compare quarterly unemployment rates for counties against zip-code level small business openings and city-level unemployment filings? To meet this challenge, we created a graph database for Hennepin County that integrates disparate data sources in ways that broaden our understanding of the impact of the pandemic across domains. Moreover, unifying our data in a graph enables us to quickly integrate new data sources. This project is the start of a promising way to integrate operational data about county services and programs, showing who the county serves and interacts with as well as the broader community impact. Our hope is that the flexibility of the graph will allow us to quickly integrate new data about county residents and their communities, improving the efficacy of government services and the well-being of our citizens. Join our session to learn more.

Speaker:

• • Lasse Andresen, CEO, IndyKite Inc

Session type: Lightning Talk

Abstract: Effective database change management relies on proper schema handling. Schemas are the "blueprint" of the database – they are instrumental for modeling and performance gains. With a strict schema, rolling out new changes or rolling back to an older version is a function of schema version control, where incremental changes are consistently tracked. This is particularly applicable for relational databases, where the schema defines the data model, type definitions, relationships, indexes, and constraints. Graph databases can be described as schema optional, or schema flexible, because there are no standard ways to manage schemas and version control in a graph database. Defining indexes and constraints in a schema is useful, but without a strict model, it's impossible to perform automatic change management. Scripts are necessary for modifying indexes and data according to the changes introduced. Database change management for graph databases, therefore, becomes increasingly challenging and is often characterized by manual efforts (and chaos), which is a hindrance to efficient CI/CD pipelines and high-performing development teams. In this talk, we'll describe in more detail how we developed a schema migration tool for graph databases that's used for automating the change management process.

Speaker:

• • Zeshan Ghory, Product Director, IQVIA

Session type: Lightning Talk

Abstract: How do we determine whether drugs are safe? Most drug regulators like the FDA and EMA maintain databases of drug adverse event reports, but these only show a small part of the picture. In this session, we will discuss how graph databases can be used to combine this data with real-world evidence (RWE) taken from hospital records, prescriptions, and insurance claims data to give a more complete picture of the safety profile of a drug.

Speaker: Donovan Bergin, Expert Software Engineer, JB Hunt

Session type: Lightning Talk

Abstract: In a world where Neo4j has solid .NET support, community drivers, and extensions, one developer asks: But what if they didn't exist and we had to get Java developers to use our library? Over the course of developing Graphr.Neo4j, we shamelessly copied from Spring Data Neo4j, learned how much reflection can damage our relationships with real-life people, and had a lot of fun along the way. And yes, the Java expats at work seem happy enough with the results.

Speakers:

• • Srinivas Kolluru, Senior Director, JB Hunt
• • Donovan Bergin, Expert Software Engineer, JB Hunt

Session type: Full Length Session

Abstract: At JB Hunt, we needed to modernize how we store, surface, and react to streaming telemetry data for well over 100,000 assets, including trucks, trailers, and containers. Neo4j AuraDB enables flexibility and performance in storing information from many disparate sources and vendors into a unified data model for use in daily operations. Moreover, it positions us to grow our graph, use cases, and capabilities as we continue our path to digital transformation. Our current architecture leverages the intake of streaming data from Kafka, writes telemetry graphs to our Neo4j instance, surfaces that data via APIs, and reacts to events with KSQL streams. We’ll present our use case, data models, and infrastructure, showing you how these technologies work in concert to provide the data and insights required to remain competitive in an ever-changing market.

Speaker: Adam Pierson, Senior Software Engineer, JupiterOne

Session type: Lightning Talk

Abstract: Starbase is an open source graph security analysis tool that collects data from external services and stores the collected data in a Neo4j database. Anyone can use it to connect to and ingest data from over 70 third-party systems into a standardized data model. Once ingested, users can perform previously complex queries quickly against the Neo4j database to gain knowledge of potential vulnerabilities. In this presentation, we will consider the value of graph-based security analysis, discuss Starbase, and briefly demonstrate a real-world use case on how data ingested by Starbase can help organizations and individuals protect themselves.

Speaker:

• • Andrea Santurbano, CTO, LARUS Business Automation

Session type: Workshop

Abstract: Apache Spark has become the most important framework for building data pipelines over the past few years because it's a framework that supports all ranges of big data formats in both batch and streaming modes. Given that, you can leverage Neo4j as a data source in Spark workflows with the official Neo4j Connector for Apache Spark. In this workshop, we'll show you how easy it is to move data back and forth in Neo4j with Spark in streaming and batch jobs using Python and Scala in a cloud environment. We'll also demonstrate how data scientists can easily combine Neo4j with Spark Python Pandas in order to provide insights.

Speakers:

• • Filippo Minutella, Chapter Lead of AI, LARUS Business Automation
• • Valerio Piccioni, AI Engineer, LARUS Business Automation

Session type: Full Length Session

Abstract: Reproducibility – both in machine learning and data science – is an emerging theme because you need to repeatedly run your algorithms on different features that can also be obtained with different graph projections to discover which one performs best on a particular dataset. Feature Store is the right tool for this objective. It can manage different versions of point-in-time features for both training and inference phases. We will present a full pipeline, starting with a graph on Neo4j and repeatedly transforming and loading it on Feast while also using Neo4j Graph Data Science to create new features to train a simple neural network. The talk will be organized with the first part on Feature Store and Neo4j's capabilities, and we will end with a notebook to present the full pipeline.

Speakers:

• • Mauro Roiter, Full Stack Developer, LARUS Business Automation srl
• • Andrea Santurbano, CTO, LARUS Business Automation Srl

Session type: Workshop

Abstract: Everybody wants secure access to data as fast as possible (near real-time), with the ability to extract meaningful insights from them. Apache Kafka and Neo4j are two of the main platforms that facilitate the achievement of this goal. Learn how you can easily integrate these two technologies and build complex streaming data pipelines that leverage the power of Kafka Connect via the Neo4j Connector for Apache Kafka. We'll show you how to setup the connector in both Source (extracting data from Neo4j and writing to a Kafka topic) and Sink (reading data from a Kafka topic and ingesting them into Neo4j) modes, and demonstrate how they work together.

Speaker: Huong Tran, Evangelist, Linkurious

Session type: Lightning Talk

Abstract: In this presentation, you'll learn tips for using Neo4j Graph Data Science to detect and fight fraud. This presentation will include information about our work with Zurich Insurance (bonus points if you take the time to read up on this client n the case study on Neo4j.com).

Speaker: Huong Tran, Evangelist, Linkurious

Session type: Lightning Talk

Abstract: Dealing with data coming from various sources? You've probably experienced the difficulties of consolidating a single view of each entity and avoiding duplicates. In this session, you'll learn concrete tips on how to apply graph analytics to tackle this challenge and power your business with accurate information.

Speakers:

• • Vineeth Venugopal, Postdoctoral Scholar, Massachusetts Institute of Technology
• • Elton Pan, Graduate Student, Massachusetts Institute of Technology

Session type: Lightning Talk

Abstract: Innovation in the materials domain is a slow and laborious process, due to which the development of new materials has been slow. A major contributing factor for this latency is the nature of knowledge organization in the sciences, where all data is unstructured and is split between different mediums. Therefore, despite decades of productive research and a profilic publication history, organized machine-readable databases are absent in the field. This is a major roadblock in the development of artifical intelligence models to enable materials prediction and discovery. In this session, you'll witness one of the largest and most comprehensive knowledge graphs in the materials domain, which is automatically extracted from a corpus of over four million published scientific articles. The knowledge graph framework is a significant develoment in the organization of materials knowledge, and through graph representation learning, is shown to not only capture complex linkages between entities, but to also discover new relations between materials, their applications, and properties.

Speaker:

• • Abbas Mohammed, Director, Data Platforms, Medifast Inc

Session type: Lightning Talk

Abstract: This session will highlight the use of Neo4j Graph Database for building a modern digital mobile app that is used by coaches of one of the fastest growing companies in the health and wellness industry. You’ll also receive an overview of the architecture and usage of Neo4j.

Speakers:

• • Mark Heckler, Principal Cloud Advocate, Java/JVM Languages, Microsoft
• • Jennifer Reif, Developer Advocate, Neo4j

Session type: Full Length Session

Abstract: Different database technologies optimize for different uses. Graph databases excel in discovering relationships, known or unknown, within vast sets of data and can help unlock value from overlooked or underutilized sources. Join the presenters in this session to discover what consideration make a dataset a candidate for graph storage and analysis. You'll also learn tips and tricks for data ingestion and structuring while gaining insights on how to build APIs that optimize for meaningful analysis of data relationships. Likewise, you'll learn how to delegate tasks to tools, automate essential but non-critical path functions, and dominate your domain with actionable insights that unlock your data's full value.

Speaker: David Meza, AIML R&D Lead, People Analytics, NASA

Session type: Full Length Session

Abstract: The beauty of knowledge graphs is the ability to expand your knowledge by connecting other domains. One way to develop these connections is through the use of natural language processing (NLP). In this presentation, we will add to NASA’s Skills knowledge graph by developing a NASA-specific skills tagger to extract entities from documents to help us find people and positions that share common skills and competencies. This approach can be used with skill mapping, skills gap, project profiles, workforce plans, and more.

Speaker: Nima Imani, Solution Engineer, Neo4j

Session type: Workshop

Abstract: This workshop will take attendees over the process of building and querying a knowledge graph of entities extracted from a set of unstructured documents (news articles) and enriched with public ontologies. We will use the APOC NLP procedures and the neosemantics plugin to import and manipulate the public ontologies. Once built, we will show how the resulting knowledge graph can be used to implement semantic search and semantic content recommendations. A prerequisite of this workshop is to have a basic understanding of the property graph model. Attendees will ideally have Neo4j Desktop downloaded and installed locally so they can code along with the presenter.

Speaker: Kristof Neys, Graph Data Science Specialist, Neo4j

Session type: Full Length Session

Abstract: Process mining is an important component for any large industrial enterprise. Recently, groundbreaking research has been performed on how graph databases are superior in analysing event logs, which has resulted in the new research area of graph-based process mining. This presentation will illustrate what graph-based process mining is, how graph data science is applied, and extend the technology to Digital Twins.

Speaker: Niels de Jong, Consulting Engineer, Neo4j

Session type: Full Length Session

Abstract: NeoDash is an open source dashboard builder for Neo4j. With just Cypher, its low-code editor enables you to visualize your Neo4j data as graphs, bar charts, tables, maps, and more. This presentation will go over how Neo4j's customers are using NeoDash to reduce time to value by making their Neo4j data visible. We will give a demo of how NeoDash can be used to build an interactive dashboard on live data in Neo4j Aura, with a variety of visualisations. We also show how a deep-link integration with Bloom lets users go smoothly from graph reporting to graph exploration, painting a complete picture of a visualization journey. Next, we review how different people in an organization can use NeoDash with Neo4j. Developers may use NeoDash to quickly prototype what a full-stack graph solution could look like, using dashboards as a tool to communicate with their stakeholders. A business user might use a dashboard to get a curated, high-level view of their graph. Ultimately, data scientists and analysts could use the different reports to analyse the results of graph algorithms, and gain a deeper understanding of their data.

Speaker: John Kennedy, Product Lead AuraDB self-serve : Free and Professional, Neo4j

Session type: Lightning Talk

Abstract: Join us for a lightning talk on Neo4j AuraDB, the graph database as a service. We'll cover all the self-serve AuraDB capabilities. AuraDB Free is a forever free tier designed to accelerate your journey from learning to production, without the hassle of managing the database. We'll cover sample datasets, loading your own data, resetting the database, and cloning or pausing your instance. You'll learn about support resources including the Neo4j Community, free training, and more.

Speaker: Dave Voutila, Sales Engineering Manager, Neo4j

Session type: Full Length Session

Abstract: In this talk, you'll learn about the Neo4j Graph Data Science team's work utilizing Apache Arrow to provide high efficiency data ingress/egress from Neo4j. We'll take attendees through an overview of the problem statement of building large graphs quickly, exporting them even faster , how Apache Arrow works, and applying Arrow in your data engineering pipeline for large-scale Neo4j use cases.

Speakers:

• • Stu Moore, Product Manager, Neo4j
• • Gustav Hedengran, Core Database Engineer
• • Tobias Johansson, Core Database Engineer
• • Linnéa Andersson, Core Database Engineer
• • Valdemar Roxling, Core Database Engineer

Session type: Full Length Session

Abstract: Join this session to hear from the experts who build Neo4j. Learn about how Neo4j is engineered for performance at scale and how its distributed cluster architecture makes it easy to scale out as needed. You may know that Neo4j is incredibly flexible and powerful, but it’s also very secure. We offer a variety of security features, including encryption, authentication, access control, and auditing. This session is for you if you're interested in learning more about Neo4j and what's coming in the next major release.

Speaker: Eric Monk, Principal Solutions Engineer, Neo4j

Session type: Lightning Talk

Abstract: Graph data modeling is an art as well as a science. It can be much more involved than the initial whiteboard model and can make a huge difference in your project. In this talk, we'll discuss the iterative data modeling process and point out some common modeling, performance, and validation issues – and their solutions.

Speaker:

• • Max Andersson, Developer Advocate, Neo4j

Session type: Workshop

Abstract: In this hands-on workshop, we'll work with the Neo4j Docker Image to set up our database instances and to monitor metrics and logs. We will go through what settings to consider and what to think about when you're setting up your own monitoring solution, as well as help you get started.

Speaker: Richard Macaskill, Product Manager, Neo4j & Costa Alexoglou, Senior Software Engineer, Neo4j

Session type: Full Length Session

Abstract: Neo4j Ops Manager (NOM) is the Neo4j solution for monitoring and administering one or more Neo4j DBMS implementations. NOM is included with Neo4j Enterprise Edition and supports single instance as well as clustered DBMSs. The initial release includes features such as a dashboard for monitoring metrics, a system topology of managed DBMSs, and a GUI for DBMS configuration and role-based access control. Attend this session to learn how NOM makes managing Neo4j operations easier and more intuitive.

Speaker:

• • Eric Monk, Principal Solutions Engineer, Neo4j

Session type: Workshop

Abstract: See how to build a React App that features a Sankey diagram using Neo4j and the GraphQL Library. Sankey diagrams are a type of visualization that show network or process flows. The talk will touch on different components of the entire stack: what data is needed for the Sankey diagram in the Neo4j web browser, how to write GraphQL and use the Neo4j GraphQL library to pull the data back, and how to make GraphQL calls to render the Sankey visualization and populate it with data. You'll see the code, libraries, and techniques to build a full-stack visualization application.

Speaker: Jeff Tallman, Sr Presales Engineer, Neo4j

Session type: Full Length Session

Abstract: In this presentation, Jeff Tallman shows how you can encrypt data stored in Neo4j using certified algorithms - from full database encryption to property-level encryption (which, in the RDBMS world, is called column encryption). He explains the impact of encryption on query performance and how industry standard requirements map to technology requirements. Jeff explains how to integrate solutions such as Protegrity and Vormetric to achieve compliance. Jeff will also describe how RBAC controls complement encryption for a full solution.

Speaker: Jeff Gagnon, Product Manager, Neo4j

Session type: Full Length Session

Abstract: In this session, Neo4j’s product manager for Bloom, Jeff Gagnon, will discuss some of the product’s new capabilities introduced over the past year, and will demonstrate how they can be used in various domains like cybersecurity, fraud detection, and more.

Speaker:

• • Jason Koo, Python Developer Advocate, Neo4j

Session type: Workshop

Abstract: Explore Neo4j’s 5.0 Python driver with Python Developer Advocate Jason Koo in this hands-on workshop. You’ll practice setting up and using the driver in local and remote scenarios, find out how to handle authentication, learn a few basic Cypher commands, and walk through resources for getting help. Bring your laptop with Python 3.7+ installed and your favorite Python editor (Jason will be using Visual Studio Code).

Speaker: Alex Woolford, Field Engineer, Neo4j

Session type: Full Length Session

Abstract: In this session we willl show how Kafka, Kafka Streams, and an ecosystem of connectors can be used with Neo4j to provide real-time insights for smarter decisions making.

Speaker: Alistair Jones, Director of Engineering, Neo4j

Session type: Full Length Session

Abstract: Neo4j Aura is a fully managed, cloud-native graph data platform, designed to be your fastest path to graph. In this talk, Neo4j Aura architect Alistair Jones will describe how Neo4j Engineering combined the best cloud technologies with innovative processes to build and operate Neo4j Aura to meet your most stringent requirements for scale, security, reliability, and ease of use.

Speaker: Jeff Tallman, Senior Presales Engineer, Neo4j

Session type: Lightning Talk

Abstract: This presentation looks at the basics of RBAC, including which privileges are available and how to construct roles to achieve the separation of duties required by many corporations and security standards. These roles include developers, data scientists, application users, and automated job processing."

Speakers:

• • Mark Andrews, Senior Professional Services Engineer, Neo4j
• • Rutvij Vyas, PS Engineer, Neo4j

Session type: Lightning Talk

Abstract: In this session, we share how we created a Python script for Levi Strauss & Co. that reads existing Cypher files, replaces the LOAD CSV commands with a process to read the file locally, and sends the values as parameters for a Cypher statement. This approach eliminates the need to host the CSV files for Aura access and in most cases, requires no changes to the Cypher files themselves. It supports Periodic Commit for sending the rows in batches and allows you to simply pass in a Cypher statement instead of requiring a file.

Speaker: Ian Pollard, Product Manager, Neo4j

Session type: Full Length Session

Abstract: User tools are evolving from separate products to the unified experience we call Workspace. In Workspace, Neo4j users can query their data, access world-class graph data visualization, and with the new Data Importer, model and import their data. Workspace combines the power of Browser, Bloom, and Data Importer into a seamless experience for Neo4j users.

Speaker: Michael Hunger, Senior Director, User Innovation, Neo4j

Session type: Full Length Session

Abstract: I was there when Cypher was invented in 2012 and have been using it ever since. The language is extremely powerful and easy to learn. But to truly master it, you need to understand how it works internally and how the database executes your queries. In this session, you'll learn to look behind the scenes at execution plans with PROFILE and EXPLAIN and which specific clauses, expressions, structures, and operations help you minimize Cypher and database operations. After this talk, you should be able to speed up your Cypher statements quite a bit.

Speaker: Marius Conjeaud, Professional Services Engineer, Neo4j

Session type: Lightning Talk

Abstract: OpenStudyBuilder is an open source project for clinical study evaluations. This tool is a new approach for working with studies that once fully implemented will drive end-to-end consistency and more efficient processes – all the way from protocol development and CRF design – to creation of datasets, analysis, reporting, submission to health authorities, and public disclosure of study information. Learn how we are building a complete solution based on a clinical knowledge graph that includes a shared API, a custom web application, and exploration tools to further analyze the data. We will also share the open source vision behind the project and how you can help! The OpenStudyBuilder, originally created by NovoNordisk, a global healthcare company, uses Neo4j as its database, along with other products from the Neo4j ecosystem, including NeoDash, Bloom, and the neomodel Python library. Neo4j's professional services team is also involved in the development and deployment of the solution.

Speaker: Stu Moore, Product Manager, Neo4j

Session type: Full Length Session

Abstract: As an organisation's data continues to grow, graph practitioners have to design solutions that enable them to make effective business decisions from multiple business graphs, or within a single graph that may be growing rapidly to multiple terabytes of data. This session is for you if you are concerned with operational issues like minimising the number of clusters, horizontal scale and elasticity, sharding very large data sets to improve manageability and querying multiple business graphs in real-time. Against the backdrop of two financial uses cases we will explore how Clustering and Fabric in Neo4j 5 - currently available as a Tech Preview in Neo4j 4.4 - can help you can make business decisions in real-time across different business graphs (federated queries) and make multi-terabyte datasets more manageable (sharding) within a highly scalable and elastic clustering architecture.

Speakers:

• • Amey Mahajan, Enterprise Presale Engineer, Neo4j
• • Alexander Fournier, Enterprise Presale Engineer, Neo4j

Session type: Full Length Session

Abstract: This presentation will be partitioned into two parts: 1) ETL best practices using the Neo4j Python driver and 2) Running supervised ML with the Neo4j Graph Data Science Library using the Python client. Part 1 (ETL): The ETL portion of the presentation will cover building Neo4j property graphs using the Python driver. We'll also go over best practices, including batching, transaction functions, templatized Cypher, and more. Part 2 (Supervised ML): The supervised ML portion will explore using the Python graph data science client. You can specify all the different properties, configurations, and user inputs that will be used to run node classification algorithms and return the results. The function calls are modular and allow the user to quickly build a graph data science pipeline and get results for any of the three node classification algorithms (fastRP, Node2Vec, GraphSage).

Speaker:

• • M. David Allen, Senior Director of Developer Relations, Neo4j

Session type: Lightning Talk

Abstract: Neo4j is unique among all databases in terms of how interactive and visual graphs are. In this lightning talk, we're going to put aside all practical work to explore and have fun with this side of graphs. We will use Cypher to create patterns in the graph to demonstrate how visualization tools help us spot patterns. Simply put, we're going to draw pretty pictures with graphs and math, and show you how you can do it too.

Speaker:

• • Dana Canzano, Support Engineer, Neo4j

Session type: Full Length Session

Abstract: This session will explore the journey of developing an in-house customer360 graph and describe the technologies employed, which include Neo4j, APOC, Structr as well as references to Zendesk, Trello, and Bambo APIs. You’ll hear about the implementation of Structr, a partner of Neo4j, along with the benefits and usage provided to users of the graph.

Speaker: David Fauth, Field Engineer, Neo4j

Session type: Workshop

Abstract: Over the past year, we released some of the most exciting and mission-critical features for the Neo4j graph database, helping developers and organizations build applications at scale with a faster time to market. In this hands-on workshop, we will take you through some of the key features of the Neo4j graph database, like relationship indexes, CALL IN Transactions, HTTP API, Single Sign-On, Helm Charts, Server-Side Routing, and more. Bonus: we will give you a sneak peek, as well as let you play with some of the experimental features that are about to come with our next major version release of Neo4j. Come and join us for some really fun and exciting hands-on labs with the product team.

Speaker:

• • Eric Monk, Principal Solutions Engineer, Neo4j

Session type: Lightning Talk

Abstract: This session will walk through how to load a JSON document in a single Cypher statement. We’ll explain how to use advanced cypher techniques such as parameterization, how to handle data quality issues, how to iterate over nested data, and conditional logic. These techniques are useful for loading JSON into Neo4j without an ETL tool or a coding language such as Python or Java; you only need Cypher. We showcase several APOC procedures and functions, as well as WITH, UNWIND/COLLECT, list expressions, and more.

Speaker:

• • Ravindranatha Anthapu, Principal Consultant, Neo4j

Session type: Lightning Talk

Abstract: As your Neo4j database grows in size, it becomes crucial to understand how to review your system for query performance and infrastructure costs. To a certain extent, the performance SLAs can be obtained by increasing memory. But once your database size becomes significant, it can be costly to keep on adding memory and CPUs. This presentation walks you through various scenarios to discuss how the growing size of your database can affect your database performance as well as how you can address those issues through data modeling and Cypher tuning.

Speaker:

• • Gerrit Meier, Software Engineer, Neo4j

Session type: Full Length Session

Abstract: Getting started with Neo4j instance and the Java ecosystem has never been easier than it is today. Be it Spring, Quarkus, or just the plain driver with the CypherDSL, get the latest update on our provided tooling and support in the Java ecosystem. We will have a drive from a basic driver example to a full object mapping supported enterprise application. In the end, you can decide for yourself what abstraction is right for you.

Speaker:

• • Richard Macaskill, Product Manager, Neo4j

Session type: Lightning Talk

Abstract: In this short session, we will cover the ins and outs of integrating a Neo4j graph database into Microsoft Excel using the Neo4j BI Connector.

Speakers:

• • Bledi Fështi, Software Engineer, Neo4j
• • Harshit Singhvi, Software Engineer, Neo4j

Session type: Lightning Talk

Abstract: In this session, we will show a demo of deploying Neo4j and Fabric via Helm charts, all in a multi-zone Kubernetes cluster. Deploying in a multi-zone cluster is a relatively complex scenario; we will show you how to use Helm charts to their full potential.

Speakers:

• • Nathalie Charbel, Software Engineer, Neo4j
• • Finbar Good, Senior Software Engineer, Neo4j

Session type: Lightning Talk

Abstract: Property graphs and property graph queries have gained tremendous traction across a wide variety of verticals and use cases. Property graph technology provides powerful tools for a variety of data analysis that could not easily be done with SQL queries against tabular data. Multiple graph database vendors competing in that space. Standards help to improve application and tool portability and developer mobility and are vital for further growth of the category. At the same time, SQL databases contain significant amounts of business critical current and historic data not readily accessible to property graph querying. The ISO/IEC and US committees responsible for standardizing SQL have recognized this trend and are actively developing two standards: 1. GQL — a full database language to allow creating, modifying and querying property graphs. 2. SQL/PGQ — an extension to SQL to present and query tabular data as Property Graphs. The Graph Pattern Matching query language syntax is common between these two efforts. This talk gives a snapshot of the current standardization efforts and timing.

Speaker:

• • Jennifer Reif, Developer Advocate, Neo4j

Session type: Full Length Session

Abstract: Java, AWS Lambda, Neo4j – one or more of these technologies might be familiar to us, but how do we use them together? In this session, we will take a look at each of these technologies by themselves, and then assemble some code to combine them. We will start with the available example code, and then see how to improve and update that code with the latest and greatest features and efficiencies offered. Through live coding, we will work through the challenges, find answers to questions, and piece together the code. As a result, we will have a working solution for a cloud function in Java that runs a query in Neo4j and returns the results. Join us to combine these technologies for a powerful trifecta!

Speakers:

• • Bledi Fështi, Software Engineer, Neo4j
• • Harshit Singhvi, Software Engineer, Neo4j

Session type: Lightning Talk

Abstract: This session will feature a live migration between the Labs Helm chart and the new Helm chart.

Speakers:

• • Brandon Campbell, Author, Ontologist and Software Engineer
• • Joel Linford, SDS Digital Innovations Lead Data Scientist, Northrop Grumman

Session type: Full Length Session

Abstract: Data fusion is a prerequisite to high-leverage analytics, but multi-source integration into data lakes becomes incomprehensible at scale. Data lakes collocate data but do not create synergy because they lack structure and context. When the time comes to engineer features, data lakes do not provide a means to maintain digital threads. The burden of preserving context falls to users, who pass tribal knowledge from one to the next through word of mouth or documentation. This process creates bottlenecks in data processing and analytics, resulting in loss of clarity over time. To overcome these challenges, Ontology-Driven Design operates on the premise that data integration should be governed by knowledge. In this paradigm, domain knowledge is modeled ontologically, which kills two birds with one stone. Firstly, the domain knowledge serves as an integration layer for disparate data. Secondly, the combination of data and ontology results in a context-rich graph that preserves domain knowledge in a digital thread. In this talk, we demonstrate how Northrop Grumman uses Neo4j graph databases to realize ODD pipelines that generate knowledge graphs can then be supercharged through analytical methods to turn data and domain knowledge into customer value.

Speaker: Andrew Flinders, Principal Data Scientist, Northrop Grumman

Session type: Full Length Session

Abstract: Free-form text often contains critical information necessary to understand a situation. However, because the user can enter text with few constraints, programmatically aggregating individual responses into a cohesive whole can be extremely difficult. Similarities between individual responses can illuminate constellations within the data that outline a bigger picture. Graph architectures are the ideal mechanism by which these connections can be revealed and explored. With the recent advent of transformer deep learning models, natural language can now be embedded into vectors that more completely capture the semantic meaning of the words. Graph analysis of similarity scores calculated between transformer embeddings provides the big picture view that is often so elusive. Thus, through a combination of deep learning, shallow learning, and graph algorithms we can extract greater insight from free-form text. In this talk, we will explore an example of this method using Neo4j and Google’s BERT transformer model.

Speakers:

• • Travis Confer, Software Engineer, Northrop Grumman
• • Steven Scott, Software Engineer, Northrop Grumman

Session type: Full Length Session

Abstract: Solving complex problems at large organizations generally requires a variety of software applications, many of which do not interface well with each other. This divides critical data across diverse data stores, hampering cross-platform analytics and understanding. One way to handle this challenge is to shuttle data between applications. However, this process results in data duplication, synchronization problems, and difficulty tracking which data source is authoritative. At Northrop Grumman, we have pioneered a graph-based solution to this problem by utilizing the GRAND Stack (GraphQL, React, Apollo, Neo4j). Our application, called Kraken, merges data from multiple sources into a graph database. This architecture simplifies master data management by decreasing data duplication, tracking changes through digital threads, and creating an ASOT (authoritative source of truth) for data. In this presentation, we will cover the story of how Kraken came to be, the problems it aims to solve, and why its graph-based architecture is useful for Master Data Management. We will also demonstrate how Kraken manages digital threads.

Speaker: Brandon Campbell, Author, Ontologist and Software Engineer

Session type: Full Length Session

Abstract: A satisfying narrative is complex and cohesive. Twisty, but intentional. Surprising and satisfying. Unfortunately, the creative process behind rich content often seems to be the mysterious byproduct of artistic genius – unapproachable without inspiration. But as a community of graph enthusiasts, we embrace complexity. We have the tools to break it down and engineer twists and turns. Usually, we apply these skills to solve business problems, but we aren't limited to them. Thinking with graphs is a superpower with applications, limited only by the creativity of the wielder. In this talk, you will see a demonstration of the art and science of designing a fictional narrative using Silky, an Electron-based React application designed for this express purpose. Silky is a graph-native word processor that encourages authors to plan their work non-linearly and unlocks the power of graph analysis through an interface to Neo4j. With it, we spin characters into threads, and threads into webs. From there, a captivating story is just a graph traversal away. This talk is for the engineer looking for a change of pace, the business professional with a new idea to model, and the hopeless authors among us with a story to tell.

Speaker:

• • David Levy, CMO, OrbitMI

Session type: Lightning Talk

Abstract: With the importance of supply chains and the global maritime trade, most assume that modern technologies make it easy for a ship owner to create a route for a crude oil tanker or cargo vessel. Wrong! Most routing services produce a default route that is the shortest distance between two ports. But the shortest distance might not be the safest (can you say pirates? storms?), or the most efficient in terms of GHG emissions. Furthermore, most routing services use unstable infrastructure, struggle with big data feeds, and offer limited visualization options. For these reasons, OrbitMI decided to build its own routing solution. To build this world-class, enterprise-grade service, we needed a graph that could process large volumes of data, provide reliable storage, serve and support spatial – in addition to linear and tabular – data sets delivered in real time (and at scale), as well as a library of pathfinding algorithms. Our goal was to develop a routing service to give ship owners and managers (really anyone in the industry), the ability to create and select the smartest route for their vessels. Routes can be optimized for distance, safety, speed, GHG emissions, weather, or revenue, among others. Learn why we selected Neo4j as a partner.

Speakers:

• • Martijn Devrome, Site Intelligence Sr. Development Specialist, Pfizer
• • Tien Do Huu, Site Intelligence Development Lead, Pfizer

Session type: Lightning Talk

Abstract: Graph databases are increasingly being used for enterprise and commercial applications. Most end-users have limited knowledge about a graph database (GDB) schema and its corresponding specific query language though. The use of a natural language interface (NLI) bridges this communication gap between the end-user and the formal query language. However, most existing NLIs are designed for relational databases and do not generalize well towards graph databases. We propose a novel and generic methodology to translate natural language questions into graph database queries. As such, end-users can type in any question related to any of the organizational data, stored in a large graph database, and immediately get the answer in both graphical and tabular format. Our framework consists of various building blocks: an intermediate query representation to uniquely and compactly store natural language queries independent of GDB query language, a semi-automatic training dataset generator, and a deep learning transformer model. By finetuning the transformer model on the generated dataset, we obtain a high prediction accuracy of 94%. In addition, given the designed diversity of the training dataset, the model is able to generalize to questions that haven’t been seen before.

Speaker:

• • Karen Passmore, CEO, Predictive UX

Session type: Lightning Talk

Abstract: Data that isn’t findable, easy to use, and actionable causes project failures, especially when you need to connect multiple, disparate sources of data and content. That’s why it’s critical for enterprises to practice design thinking to ideate, design, and test graph-based products when collaborating with users, data consumers, business leaders, and data experts. Design thinking is a UX method for collaboratively and iteratively designing content and data solutions based on user needs and testing. It puts business, data experts, data consumers, and users at the center by defining user journeys and pain points on both the front-end and backend to tie app and data flow to scenario-specific outcomes. This approach puts companies on track to reap their expected returns and nurture the health of content and data products with a defined governance structure for managing change over time. The result: apps full of graph-powered data insights and integrated, usable experiences that grow and protect revenue. In this talk, we take a look at design thinking applied to graph data, using real-world case studies from our clients.

Speaker:

• • Priyanka Angadi, Data Science Manager, PwC

Session type: Lightning Talk

Abstract: "Synthetic graphs offer a powerful method for implementing privacy by design. We implemented several research-based algorithms on real-world data. In particular, we focused on two algorithms: synthetic graphs by Mittal et al. that adds nodes and edges to a graph while persevering their utility; and the uncertainty injection approach of Boldi et al. that introduces edge probabilities to obfuscate graph information. Although tested on research data, these algorithms posed certain challenges when using them on real-world data. This presentation will cover lessons learned and the modifications we applied when processing data with personally-identifiable information."

Speakers:

• • Nouamane Bensaoud, Senior Consultant, Sopra Steria
• • Daniel Schmitz, Data Science Consultant, Sopra Steria

Session type: Full Length Session

Abstract: Telecom providers suffer from outdated, manually-maintained inventory data, which leads to network mismanagement and impedes network automatization. To overcome these challenges, Sopra Steria developed a solution that reconstructs the actual network topology from network configuration data, stores it in a Neo4j database, and visualizes the result in a comprehensible way: the Intelligent Network Analyzer. We show how Neo4j can be employed to efficiently store and analyze a country-wide telecom network. Exemplary use cases are tracking the evolution of the network over time while providing impact analysis and fine-grained network exploration. Furthermore, we demonstrate how to perform a traffic simulation using Neo4j. Based on Neo4j and networkx, an algorithm optimizes the traffic routing in the network to minimize overloads and latency. The presentation includes a live demo.

Speaker:

• • Henrik Johansson, CTO, Stratazon AB

Session type: Full Length Session

Abstract: In a world of global expansion and digitalization, businesses have never been more keen to understand their customers. The most trustworthy, untouched, and up-to-date data pipeline to capture that understanding is a currency of the future – customer reviews. By tapping into the stream, businesses can gain access to an ever-growing flood of consumer feedback. But without context, ratings and reviews are merely digits and characters. With graph technology, you can integrate product and brand reviews with transactional data in a tactical context to create tangible and actionable insights. This ultimately facilitates rapid decision-making to create real-time business value.

Speaker: Jayshree Kottapalli, Industry Advisor, TCS

Session type: Full Length Session

Abstract: With the Telecoms set to roll out 5G offerings, and world on the brink of adopting 5G network, it becomes imperative for the Telecoms to gear up their supporting system to process, analyze and action the voluminous and fast data that the new age promises to bring - be it to provide proactive and predictive Network service assurance, IT supervision, Network security against cybersecurity threats, software driven Network & Data center management, IoT functions etc. It is with graph analytics anchored on graph databases that enables this to be implemented effectively. Graph data platforms provide Telcos with the massive flexibility of connecting data from multiple siloed systems to bring together a hyperpersonal service to its customers. We will take a quick view of how graph systems and analytics are ideal to accompany Telcos and anchor them in this journey into the world.

Speaker:

• • Kasthuri Kannan, Associate Professor, University of Texas MD Anderson Cancer Center

Session type: Full Length Session

Abstract: For decades, relational databases and static networks have been the mainstream for data-driven biology and medicine. Despite the simplicity of these approaches, these static architectures do not enable unbiased hypothesis generation and systems biology-based discovery. Systems biology attempts to capture the dynamic and integrative nature of biology. Graph databases as mutable and dynamic querying structures coalesce data through interactions, enabling the study of biological systems as complex adaptive systems and providing a platform for hypothesis generation. This article introduces a generalized patient-centric graph schema that integrates molecular and clinical datasets, providing an unbiased hypothesis-generation paradigm for cancer research. It also highlights an example of a database analysis using a commercial graph solution (Neo4j) that reveals the integrated landscape for a molecular subtype of brain tumors.

Speaker:

• • Chase Clark, Postdoctoral Research Associate, University of Wisconsin-Madison

Session type: Full Length Session

Abstract: Genomics-based bacterial drug discovery is heavily focused on searching for groups of genes that encode for the production of medicinally-useful chemical compounds (e.g. antibiotics, anti-cancer drugs, etc). Finding genes/groups of genes related to those that encode for known chemicals allows us to find genes that encode the production of similar compounds that may have better activity or even completely different clinical uses. However, current tools are limited in their ability to scale past dozens of genomes and/or rely on precalculating predicted groups of genes. In this session, you will learn how I created Socialgene using a unified collection of tools (Neo4j graph database, Python package, Django app, Nextflow pipeline) to enable rapid, flexible searching through large numbers of genes and genomes. While still in the early stages of development, Socialgene has been able to discover related genes and groups of genes in hundreds to thousands of genomes in near real time. My current work is focused on developing a reproducible and robust data-generation, database-creation pipeline, and future plans include creating a repository-scale database through use of the University of Wisconsin-Madison’s Center for High Throughput Computing.

Speaker:

• • Leen Schafer, Chief of Staff, Viasat

Session type: Full Length Session

Abstract: Viasat's mission is to connect the world. Our rapid growth in response to the needs of a global market presents new challenges around nurturing intracompany connection, collaboration, and communication. To mitigate these challenges, our intranet, Compass, leverages graph, a robust search engine and machine learning tool to create a best-in-class employee experience. As an enabling team at Viasat, we seek out new and innovative ways we can use our data to help us become a more connected company.

Speaker: Khushbu Agarwal, Senior Research Scientist, Pacific Northwest National Laboratory

Abstract: The need for automated assistants for healthcare is well-established. Medical providers benefit from an automated assistant who helps to triage patients, summarize changes in a patient’s condition, and suggests possible drugs or procedures. Similar needs arise in home healthcare, where automated assistants connect caregivers and physicians to monitor the patient’s health and review treatment options. Chatbots are seen as a promising technology for such mediation. However, such systems lack the reasoning capabilities required to explain the rationale behind a recommendation and to communicate with the end user. Medical AI systems must understand the relationships between symptoms, medical conditions, drugs, and procedures. These relationships are most naturally represented as a knowledge graph. In this session, we will demonstrate a neural-symbolic reasoning system that builds a symbolic representation of clinical knowledge from multi-modal electronic healthcare records data using a combination of scalable graph data mining and deep learning. We will showcase a natural language interface that allows an end user to interact with the system to identify high-risk patients, review summaries and recommendations, and improve the knowledge graph and models by answering questions.

Speakers:

• • Luis Eduardo Almazán, Data Science Consultant, VinkOS
• • Chris Upkes, Principal Professional Services Consultant, Neo4j

Session type: Full Length Session

Abstract: Neo4j is greatly settled with a good presence in some countries, but it can be an uphill journey in some other ones. In this session you'll learn about the graph journey to implement a solution in a market that doesn’t know much about graph databases.The immersion into graph technology and methodology doesn't always translate immediately to value for clients – there's exploration to be made, discoveries to be found, skills to be developed, and a learning curve for everyone involved. We'll go over challenges and struggles, how to overcome them, and how to achieve a successful journey with graphs. We are going to tackle similar problems with similar customers, except in this case, our environment will be completely unfamiliar. Over 40 minutes, we are going to walk along with VinkOS on their journey to solve some of the biggest fraud analytics problems with their Fortune 500 customer, as they work alongside Neo4j to identify essential requirements, design and test an initial solution, and work with the customer to evaluate the solution and determine a working relationship. Unlike other use-case specific demonstrations, in this presentation, you will get an intimate understanding of what it takes for Neo4j’s first Latin American partner to identify, design, and solve a fraud detection use case for a Fortune 500 customer.

Speakers:

• • David A Stumpf, Principle, Who Am I, LLC
• • Weidong Yang, Founder & President, Kineviz

Session type: Full Length Session

Abstract: Graphs for Genealogists (GFG) is an open-source software package with an application front-end, visualization in GraphXR, a graph database, and a plugin designed for genealogy data management and analytics. The ETL loads family tree data in GECOM format, consumer DNA test results, and genealogist-curated files that create links between graphs. The analytics provide new insights and actionable recommendations for further genealogy research. GFG traversals collect concatenated strings to create Ahnentafel numbers and enable filtering on X-linked inheritance and other patterns. Traversals from the family tree through DNA matches to chromosome segment data find triangulation groups and monophyletic segments aligned with specific family tree branches. Graph algorithms from Neo4j Graph Data Science reveal communities aligned with family tree branches. Hierarchical trees include patrilineal trees, DNA haplotrees, ORDPATH-enhanced renderings, and hybrids linking these together. Chromosome painting and 3D renderings help users interpret the results. Recommendations include manageable sets of persons from a pool of over 250,000 DNA matches. There are many opportunities for further development of graph analytics in this billion-dollar industry.