Chronon
chronon.ai
What it is
Chronon is an open-source feature platform for building, deploying, managing, and monitoring machine-learning data pipelines. It computes features in batch and in real time for online serving, offline training, or evaluation. Features are defined through a Python API with transformations and aggregations, drawing on inputs such as event streams, warehouse tables, production databases, change data streams, service endpoints, and request-context data. Point-in-time-correct backfills are intended to keep offline training features aligned with production inference data. Chronon can create monitoring pipelines for training-data quality, training-serving skew, and feature drift. Batch processing uses 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 data integration guide assumes access to a Spark environment connected to a warehouse. The project is self-hosted and published under the Apache 2.0 license. Its Open source plan is free, with no usage limits stated.
Spec sheet
- What it does
- Chronon is an open-source feature platform for building, deploying, managing, and monitoring machine-learning data pipelines.Read 4 Oct 2026 · docs.chronon.ai
- Product
- Chronon is an open-source feature platform for building, deploying, managing, and monitoring machine-learning data pipelines.Read 4 Oct 2026 · chronon.ai
- Support
- The project README directs users to GitHub Issues for bugs or feature requests and its community Slack for discussion, tips, and support.Read 4 Oct 2026 · github.com
- License
- The Chronon project is published under the Apache 2.0 license.Read 4 Oct 2026 · github.com
- Adoption
- The project lists Airbnb, Stripe, OpenAI, Roku, Netflix, Uber, Intuit, Airwallex, Sardine AI, and Monzo Bank as adopters.Read 4 Oct 2026 · github.com
- Batch and streaming
- Chronon uses Spark for batch jobs and supports streaming with Spark Streaming or an experimental Flink connector.Read 4 Oct 2026 · docs.chronon.ai
- Compute integrations
- The homepage describes Apache Spark for batch processing and Apache Flink for streaming aggregations.Read 4 Oct 2026 · chronon.ai
- Consistency
- Chronon supports temporally accurate offline backfills intended to keep training data consistent with production inference data.Read 4 Oct 2026 · chronon.ai
- Data inputs
- Documented inputs include event streams, warehouse tables, production databases, change data streams, service endpoints, and request-context data.Read 4 Oct 2026 · chronon.ai
- Data sources
- Chronon can consume event streams, database snapshots, change data streams, service endpoints, and warehouse tables.Read 4 Oct 2026 · docs.chronon.ai
- Deployment requirement
- Chronon jobs require Spark, and users configure execution through a spark_submit.sh script.Read 4 Oct 2026 · docs.chronon.ai
- Feature computation
- Chronon supports batch and real-time feature computation for online serving and offline training or evaluation.Read 4 Oct 2026 · chronon.ai
- Feature definitions
- Users define features with a Python API using concepts such as GroupBy, Join, and StagingQuery.Read 4 Oct 2026 · chronon.ai
- Feature monitoring
- Chronon can generate monitoring pipelines for training-data quality, training-serving skew, and feature drift.Read 4 Oct 2026 · docs.chronon.ai
- Getting started
- The project quickstart requires Docker and describes running Chronon locally with a docker-compose file.Read 4 Oct 2026 · github.com
- Monitoring
- It can generate monitoring pipelines for training-data quality, training-serving skew, and feature drift.Read 4 Oct 2026 · chronon.ai
- Orchestration
- Airflow is the best-supported orchestration method, though the documentation says other integrations are possible.Read 4 Oct 2026 · docs.chronon.ai
- Point-in-time correctness
- Chronon supports point-in-time-correct training backfills to help keep offline training features consistent with online serving.Read 4 Oct 2026 · docs.chronon.ai
- Security reporting
- The repository security policy asks users to report potential security issues through Airbnb's Bug Bounty program rather than public issues or pull requests.Read 4 Oct 2026 · github.com
- Serving API
- Chronon provides Java and Scala Fetcher libraries, and the Fetcher can also be used to create a service callable over REST.Read 4 Oct 2026 · docs.chronon.ai
- Setup requirement
- Chronon jobs require Spark to run, and the data integration guide assumes access to a Spark environment connected to the data warehouse.Read 4 Oct 2026 · docs.chronon.ai
- Storage integration
- Online serving requires integrating a key-value store that supports point writes, point lookups, scans, and bulk writes.Read 4 Oct 2026 · chronon.ai
- Streaming integration
- For online streaming, Chronon requires a StreamDecoder to convert Kafka message bytes into Chronon events or mutations.Read 4 Oct 2026 · chronon.ai
Plans and prices
Strong points and limits
Strong points
- Supports batch and real-time feature computation
- Python API defines transformations and aggregations
- Point-in-time-correct backfills support training consistency
- Generates pipelines for feature and data monitoring
- Published under the Apache 2.0 license
Limits
- Jobs require Spark to run
- Online serving needs a user-provided key-value store
- Flink connector is described as experimental
Who it suits
Chronon suits teams building machine-learning pipelines that need batch and streaming feature computation, offline backfills, and online serving. Users need a Spark environment and must provide a key-value store for online serving.
Verdict
Chronon covers feature computation, serving, backfills, and monitoring in a self-hosted platform. Its Spark requirement and infrastructure integrations are important setup considerations.
Where Specifiction places it
- Feature Store Software#1 of 208.4