Graph Database Stories: Google and Netflix

Graph databases are databases optimized for storing and querying highly-interconnected data. They leverage graph structures for semantic queries with nodes (also called vertices or points), edges, and properties to represent and store the data. A key concept of the system is the graph (or edge or relationship), which directly relates data items in the store. The relationships allow data in the store to be linked together directly, and in many cases retrieved with one operation.

Join us to learn how Google created a temporal graph database and how Netflix deployed TitanDB at scale.

Schedule:

6:15 - 7:00: Refreshments and Networking
7:00 - 8:30: Talks
8:30 - 9:30: Refreshments and Drinks

BadWolf @ Google

BadWolf is a temporal graph store abstraction loosely modeled after the concepts introduced by the Resource Description Framework (RDF). It provides a flexible storage abstraction, efficient query language, and data model for representing temporal directed graph that accommodates the storage and linking of arbitrary objects without the need for a rigid schema. BadWolf's main focus is around temporal graph, or graph that contains facts that change over time. BadWolf began as a triple store, but triples soon got expanded to incorporate time as a first class citizen to allow simpler and flexible temporal reasoning. This talk will cover BadWolf basic concepts, show how data can be modeled using such concepts, and demo some of its temporal reasoning components.

Speaker Bio:

Xavier Llorà is the tech lead for Google's Counter-Abuse Technologies team. His work focuses on helping fight abuse across Google properties at scale to keep users safe online. He is also the lead for the OSS effort named BadWolf ( https://github.com/google/badwolf ). Xavier joined Google after working for nine years as a research assistant professor at the National Center for Supercomputing Applications and the University of Illinois at Urbana-Champaign. There, he combined research on scalable genetics-based machine learning, supercomputing infrastructure, and knowledge representation for automated innovation.

TitanDB @ Netflix

Content Platform Engineering at Netflix develops technology and operations for most aspects of streaming content that make Netflix a great global service. The platform for storing the vast and growing amount of content metadata is stored in a graph database utilizing TitanDB. The platform exposes semantic APIs for the users to be able to store, tag, search and retrieve metadata. In this talk, we will present the high-level graph model that we use to store these metadata, the learnings we have had from scaling the system to support different use cases and over 1 Billion vertices.

Speaker Bios:

Firouzeh Jalilian is part of Content Platform Engineering Infrastructure team at Netflix focusing on digital asset management. Before this, she was in the Account and Partner Platform team at Netflix developing partner payment integrations.

Vikram Singh is a software Architect at Netflix. Currently a member of the content infrastructure engineering team, before that worked wth the Membership engineering team. At Netflix focused on building distributed scalable architectures. Co-author of Netflix Conductor, an open source orchestration engine for microservices. Worked on Digital Asset Management system which is based on Graph DB

Parking Map:

The event will be in Building D

https://secure.meetupstatic.com/photos/event/b/5/2/f/600_464326383.jpeg