Getting Connected at AWS re:Invent

Getting Connected at AWS re:Invent

TigerGraph’s product management team and I attended AWS re:Invent, where 60,000 persons from around the globe had gathered to see what is new and exciting in the world of cloud services. With literally thousands of sessions and events to choose from, I could only get a taste of all that is happening on the cloud.  

A few themes stood out:

Machine Learning and AI are ubiquitous. From controlling autonomous vehicles to making a better cup of coffee, machine learning is being used in the full range of applications. A product manager for a team that demonstrated graph-based neural networks confided in me that most of the graph databases out there are not performant enough to do machine learning in-graph; the data has to be exported to something like TensorFlow or AWS SageMaker and run there. His data scientist said there’s only one graph DB that they know of that would be performant enough. He checked with his scientist to get the vendor’s name: TigerGraph.

Graph is growing. There were dozens of talks which included graph databases or graph analytics. The number of questions from the audience showed a deep and serious level of interest. Graph speeds up multi-join queries, and it enables data modeling and querying centered around how things relate to one another. The TigerGraph Cloud service is making graph available to everyone.

Containerization and scalability of service are vital. Containers reduce a lot of the stress of deployment and maintenance for system administrators, even on the cloud. Organizations also want to know that a service can grow as their needs grow. TigerGraph hears you. You can run TigerGraph in Docker. We plan to offer full Kubernetes support next year. Next week, the TigerGraph Distributed Cloud will be available, allowing you to deploy a distributed graph database, the only such SaaS service that scales both storage and compute across multiple instances.

It’s about interconnection. Whether that’s using a graph database to merge your data silos into one master dataset, or running GraphQL federated queries across multiple data sources, running graph analytics and machine learning to gain insight from seeing how things interrelate,  or building a full or hybrid cloud data pipeline, the need to connect one service to another is foundational. TigerGraph has had REST APIs and JSON output from Day 1, for universal connectivity. We’ve since added many other targeted connectors, drivers, and APIs: JDBC, S3, Kafka, Python… Enabling you to connect TigerGraph to the service of your choice is a major priority for us in 2020.

2019 has been a great year for TigerGraph. From what I saw at AWS re:Invent, 2020 is going to be even better.

We encourage you to try TigerGraph. One of the best ways to do that is with TigerGraph Cloud. It takes only 10 minutes to get started and it’s free. It’s also easy to get started. TigerGraph Cloud contains 15 starter kits for uses cases such as customer 360, data lineage, entity resolution, explainable AI, fraud detection, machine learning, social network analysis and supply chain management. These starter kits contain pre-populated graph schema and queries – so you could be building your very own proof-of-concept in just a few hours. And deploy it to production in days. You can subscribe at

You can deepen your knowledge of graph by reviewing the many educational resources on our website. One of the best of these is our TigerGraph certification program. Becoming a certified TigerGraph associate is an important first step on the curve to learning more about graph database and analytics. Not only will this provide you a good understanding of the business benefits and use case cases of graph, it will also provide you the basics of programming in GSQL. This course consists of a series of informative and interesting videos and a test. We’ll send you a certificate once you pass the test along with a t-shirt, if you request one. You can begin your certification at