The Incremental Compute Engine for AI, ML and data teams

Feldera is a fast query engine with the unmatched ability to evaluate arbitrary SQL programs incrementally, over any number of live and historical data sources. Process millions of events per second and scale to larger than memory datasets, even on your laptop.

Use cases

Real-time Feature Engineering

Get perfect offline-online parity, zero feature skew, instant feature freshness and process millions of events per second. Ship new features in days (not months), by reusing the same complex SQL to do feature engineering in both offline (batch) or online (real-time) settings.

    alt text

    Use cases

    Threat, identity and fraud platforms

    Run thousands of complex queries to incrementally enrich, correlate and aggregate over multiple signals in real-time to stop bad actors. Whether ML-based or rule-based, compute with perfect offline-online parity at millions of events per second.

      alt text

      Use cases

      Operational analytics

      Incrementally compute over unbounded streams of live and historical data generated from your cloud and edge. Integrate-with and alleviate pressure from your lakehouse or warehouse.

        alt text

        VMware Skyline used the Feldera team's technology to digest terabytes of streaming data and execute thousands of complex rules instantaneously. We reduced our end-to-end recommendation notification time from twenty-four hours to minutes. The engine is ridiculously fast and was never a compute bottleneck for the three years it has been in production.

        Portrait of Alex Bewley

        Alex Bewley

        Director, Engineering - VMware

        Incremental Computation has never been this easy

        Go from painful data-engineering and ad-hoc solutions to “Just Writing SQL”. Define tables and deeply nested layers of views. When inputs receive changes (inserts, updates or deletes), all views are incrementally updated in an instant. Try it right away.

        Incremental Computation has never been this easy

        The state-of-the-art in Incremental Computation

        Powered by our open-source compute engine, written from scratch in Rust.

        Star us on GitHub
        A unified query engine for offline and online compute, no compromises

        A unified query engine for offline and online compute, no compromises

        Evaluate complex SQL over both batch and streaming data sources, seamlessly handling inserts, updates and deletes.

        Documentation
        Unmatched performance

        Unmatched performance

        Our users routinely see millions of events per second on a single machine, out-of-the-box, even for sophisticated pipelines with complex views computed over many streams.

        Documentation
        Fully automatic incremental compute, powered by award-winning research

        Fully automatic incremental compute, powered by award-winning research

        We’ll take care of the incremental algorithms, strong consistency, spilling to fast storage, failure recovery and keeping memory stable even when computing over infinite streams.

        Documentation

        Feldera pipelines can connect to multiple heterogeneous data sources and destinations.

        Logo 1Logo 2Logo 3Logo 4Logo 5Logo 6Logo 7Logo 8Logo 9Logo 10

        The Feldera Community

        Join our growing community of users, engineers and enthusiasts to discuss incremental computation, databases, streaming, Rust, and all things data.

        Star us on GitHub

        Join our Slack or Discord communities