Standalone or embedded

Streaming SQL
in one process

Run SQL over Kafka events. Count usage, join streams and maintain live aggregates. Deploy one binary or embed the engine in your application.

Try it liveInstall Ubik

No install or signup for the demo. Free evaluation; paid commercial licences.

Orders per merchant
ubik --from kafka://demo.getubik.dev/orders \  "SELECT merchant,          TUMBLE(ts, INTERVAL '5' SECOND) AS w,          count(*) AS orders   FROM orders   GROUP BY merchant, TUMBLE(ts, INTERVAL '5' SECOND)"

Each completed window produces a count per merchant.

Run this query

v1.3.0Recovery & guarantees Benchmarks

From events to results

Read your stream

Connect to your existing Kafka broker. Use a local JSON file for a replay.

Write the query

Filter events, aggregate by customer or time window, and join related streams.

Run the pipeline

Write results to Kafka or consume them in your application. Checkpoint state and offsets for recovery.

One process per pipeline. Deploy it on your infrastructure or embed it in your application.

Get started
Live shellThe released binary in your browser, on a live stream. No install.Downloadbrew, pip, or one checksum-verified tarball, and a durable pipeline in one command.ProductSQL, sources, joins, windows and deployment options.