Conference Graph + AI World 2020 – graph algorithms and machine learning

Graph + AI World

The conference will be held on September 28-30 Graph + AI World 2020 for people who are not indifferent to graph technologies and machine learning. The event will be held online for three days, participation is free.

The organizer was the company TigerGraph, the creator of the eponymous Grafova DB, and the program will include reports from speakers from various companies: Intel, KPMG, AT&T, Forbes, Intuit, UnitedHealth Group, Jaguar Land Rover, Xilinx, Xandr, Futurist Academy, etc.

Why participate as a Leader or Engineer and join one of 3,000 members from 110 Fortune 500 companies? Welcome to cat.

For those who want to take part right away, a link to registration

The Graph + AI World conference aims to improve the efficiency of AI and machine learning projects through the use of Graph Algorithms.

Why Graph Algorithms?


We use graph databases every day and are probably not aware of it. Facebook, Instagram and Twitter use graph databases and analytics to understand how users relate to each other and link them to the right content. Every time you do a Google search, you are using the knowledge graph from Google. Product recommendations on Amazon – “people who bought this product also bought …” or “these products are often bought together”? All this is also associated with analytical queries to graph databases.

If we compare different types of databases, we can highlight the main trends:



graph databases

Relational databases

Complex, slow, need to link tables

  • Rigidly built scheme;
  • High performance for transactions;
  • Poor performance for deep analytics.

Key-value database

Requires multiple scans of an array of tables

  • There is no clear outline;
  • High performance for simple transactions;
  • Poor performance for deep analytics.

Graph databases

Preconnected business entities – no need to link objects.

  • Flexible scheme;
  • High performance for complex transactions;
  • High performance for deep analytics.

Thus, if your data has many relationships with each other, it is logical to use Graph databases instead of multiple Join queries, which will not be so effective on large volumes. Besides, nobody canceled Graph Theory for Data Science;)

Key speakers

Graph + AI World 2020 Key Speakers

  • UnitedHealth Group created the largest Graph Database in the healthcare industry to communicate, analyze, and provide real-time advice on treatment pathways for 50 million patients.
  • Jaguar Land Rover have reduced request times for their complex supply chain model from 3 weeks to 45 minutes, allowing them to plan accurately and quickly respond to supply and demand uncertainties in the wake of the Covid-19 pandemic.
  • Intuit use knowledge graph as fundamental technology for AI driven expert platform.


The conference has a stellar agenda, filled with training and certification sessions on September 28 (preliminary day) and business cases, use cases, and technical sessions on September 29 and 30. Some of the sessions are highlighted below.

28 September

Introduction to Graph Algorithms for Machine Learning Certification

Graph algorithms are essential building blocks for related data analysis and machine learning to gain a deeper understanding of that data. Graph algorithms can be used directly for unsupervised learning or for enriching training samples for supervised learning. This lesson will introduce the new TigerGraph training and certification program for applying Graph Algorithms to Machine Learning: content review, video, demo and certification process.

Hands-on Workshop: Accelerating Machine Learning with Graph Algorithms

In this workshop, you will be able to apply several different approaches to machine learning with graph-based data.

After setting up your graph database (in the cloud and free), we will do the following:

  • Unsupervised Learning Using Graph Algorithms
  • Feature Extraction and Graph Enrichment
  • External training and integration with notebooks
  • In-database ML techniques for graphs

We will have several datasets for different cases.

September 29

Application of Graph Model in Fintech and Risk Management

FinTell has built a graph with tens of billions of edges and nodes based on 1.5 billion active mobile devices per month. The graph model helps FinTell deliver superior quality risk management services to financial institutions.

Building a State of the Art Fraud Detection System with Graph + AI

A step-by-step guide and demonstration of what analytics can be quickly built using graph analytics on modest computing resources, and how anti-fraud metrics are improved by reducing missed fraud cases AND reducing false positives in a standard machine learning pipeline.

Executive Roundtable – Transforming Media & Entertainment With Graph + AI

Graph databases are used to identify, link and combine repeating customer entities and to build an insightful 360 ° view of customers. This usually translates into higher returns as a result of more accurate and effective recommendations for products and services. Join the executives at Ippen Digital and Xandr (part of AT&T) to learn how graphs and machine learning are changing the media and entertainment landscape.

September 30th

Supply Chain & Logistics Management with Graph DB & AI

Industrial manufacturing faces significant challenges with the sheer number of parts, components and materials that must be purchased from a multitude of globally distributed suppliers and then processed and assembled at multiple stages, making it difficult to trace from supplier to final product. This also includes logistics, i.e. types of transport, locations, duration, cost, etc.
By leveraging Graph Databases to provide transparency for complex and distributed data, coupled with predictive analytics, manufacturers can effectively address these challenges. Simultaneously optimizing production planning: ensuring parts availability, minimizing quality loss, improving assembly and overall delivery.

Recommendation Engine with In-Database Machine Learning

Recommender systems are used in various services such as video streaming, online shopping, and social media. In industrial applications, a database can contain hundreds of millions of users and items. Training the model in the database also avoids exporting the graph data from the DBMS to other machine learning platforms, and thus better maintain a continuous update of the recommendation model on changing training data.

Also at the conference the results of the hackathon will be summed up Graphathon 2020


To participate in the free conference, you must register on the official page of the event. link

Join Graph + AI World!

See you at the conference)

Similar Posts

Leave a Reply