Abstract
Chronon is a data processing framework open-sourced by Airbnb. It is adopted across organizations like Stripe, Netflix, OpenAI, and Uber. Chronon was originally built for ML applications. It has since been adopted to power a variety of use-cases—heuristics for rule engines, context for LLMs, user-facing and business-facing metrics.
Chronon is adopted for its ability to generate training data at scale and serve features with very low latency with a simple, high-level API. It abstracts away the effort required to manually build batch and stream processing pipelines, indexes, and services.
This talk will focus largely on algorithms and optimizations inside Chronon. We will only briefly touch upon the core concepts of the API and a couple of example use-cases.
Speaker
Nikhil Simha
Co-Founder & CTO @zipline.ai, Author of "Chronon Feature Platform", Previously @Airbnb, @Meta, and @Walmartlabs
Nikhil Simha Raprolu is the Co-founder & CTO at zipline.ai. Prior to that he worked on the ML Infra team at Airbnb - where he open-sourced Chronon. At Facebook he worked on stream processing systems, schedulers and compilers - eg., Stylus & Turbine. Prior to that he worked on distributed data processing infrastructure at Amazon and Walmart Labs.
Find Nikhil Simha at:
Date
Tuesday Mar 17 / 02:45PM GMT ( 50 minutes )
Location
Fleming (3rd Fl.)
