Chronon is an open-source feature platform for building, deploying, managing, and monitoring machine-learning data pipelines. It computes features in batches and in real time for online serving and offline training or evaluation. Teams define features through a Python API with transformations and aggregations, drawing on inputs such as event streams, warehouse tables, production databases, service endpoints, and change data streams. Point-in-time-correct backfills are intended to keep offline training features consistent with online serving. Chronon can generate monitoring pipelines for training-data quality, training-serving skew, and feature drift. Batch jobs use Spark; streaming options include Spark Streaming and an experimental Flink connector. Online serving uses Java or Scala Fetcher libraries embedded in a service or application, with a user-provided key-value store such as Redis or DynamoDB. Jobs require Spark, and the quickstart describes local setup with Docker and docker-compose. Chronon is self-hosted and published under the Apache 2.0 license. Its open-source plan is 0.00 USD per free, with no usage limits stated.
Who it is for
Chronon suits machine-learning teams building and operating feature pipelines who can run Spark jobs and self-host the platform. Its Python feature definitions and online and offline workflows address both serving and training needs.
What is good
- Supports batch and real-time feature computation
- Provides point-in-time-correct training backfills
- Can generate feature and data quality monitoring pipelines
- Published under the Apache 2.0 license
What to know first
- Chronon jobs require Spark
- Online serving requires a key-value store
- Quickstart setup uses Docker and docker-compose
- Flink connector is experimental
Verdict
Chronon covers feature computation, online serving, offline training, and monitoring in a self-hosted open-source platform. Teams should account for the Spark requirement and the supporting infrastructure needed for serving.
Chronon plans and pricing
All plansCompared on feature store software
- Online store
- Yes
- Offline store
- Yes
- Point-in-time joins
- Yes
- Feature monitoring
- Yes
- Deployment model
- self_hosted
- Serving modes
- both
