MLOps.community Real time features, AI search, Agentic similarities
16 snips
Dec 28, 2025 Varant Zanoyan, Co-founder and CEO of Zipline AI, and Nikhil Simha Raprolu, Co-founder and CTO at Zipline AI, delve into the evolution of AI infrastructure. They share insights on the compute-first approach of Cronon that emerged from Airbnb, emphasizing real-time features over traditional storage models. The duo explains the complexities of orchestrating signals and pipelines, the challenges of point-in-time correctness, and the importance of governance. They also discuss how Cronon integrates embeddings and real-time workflows, reflecting on its open-sourcing journey with Stripe.
AI Snips
Chapters
Transcript
Episode notes
Feature Platforms Must Be Compute-First
- Chronon treats feature platforms as compute-first systems that produce, serve, and sync online and offline features reliably.
- This compute-centric view solves streaming aggregations, windowing, and point-in-time correctness that many stores ignore.
Cronon Born From Fighting Payments Fraud
- At Airbnb the early Cronon versions were built to fight payments fraud and enable rapid iteration for adversarial use cases.
- Teams needed to engineer features in hours because fraud patterns required very fast responses.
Models Need Thousands Of Signals
- Production ML pipelines require hundreds to thousands of simple