Real-time Analytics in the Cloud: Running Real-time A/B Experiments at Massive Scale

Real-time data is incredibly valuable and actionable but organizations still face obstacles in their path to real-time analytics. Instant personalization, fraud fighting, support operations, large scale A/B testing— all have requirements beyond what a traditional or cloud data stack can provide. In this talk, Rockset Co-founder and CEO Venkat Venkataramani will show how a real-time A/B experiments system can be built at massive scale in the cloud.

Based on experience from scaling real-time A/B experiments systems at facebook, we find that real-time A/B experiments platforms ideally have the following requirements:

• Speed with scalable real-time ingest: This will help product teams make decisions in days instead of weeks. The faster results arrive, the more experiments product teams will run.
• Ability to analyze multi-dimensional data: Almost every part of A/B testing analysis involves JOINing the real-time event stream with one or more fact tables such as users, products, devices or experiments data.
• Sub-second queries with interactive slicing & dicing: Product teams need to drill-down and interrogate the data in an interactive fashion to build new hypotheses, constructs better ideas and design follow-up experiments.

Venkat will introduce Rockset, a real-time indexing database in the cloud, and show how its schemaless ingestion, full-featured SQL, and cloud scalability can be used to build a real-time A/B experiments system that meets these requirements.