JFrog ML Feature Store
qwak.com
- Free trial
- Linux
- Web
- API
- Self-hosted
What JFrog ML Feature Store is
JFrog ML Feature Store centralizes the development, sharing, and serving of machine-learning features to help keep training and inference data consistent. Teams can define features once and reuse them across projects, retrieving historical values for training and online values for inference. Feature sets may be batch, streaming, or real-time, but JFrog ML SaaS supports batch feature sets only; streaming and real-time sets require hybrid deployment. Listed batch sources include Snowflake, BigQuery, MongoDB, Amazon S3 files, Redshift, MySQL, Postgres, ClickHouse, Vertica, and AWS Athena. Kafka and Kinesis are listed as streaming sources. Users can set up data sources and feature sets with the SDK or CLI, or create data sources in the web interface. JFrog ML supports AWS, Google Cloud, and Microsoft Azure, with deployment on JFrog's platform or customer infrastructure. Models can be served as live API endpoints, batch inference jobs, or Kafka stream applications. The product is presented for data scientists, ML engineers, and AI developers moving AI and ML services into production.
Who it suits
It suits data scientists, ML engineers, and AI developers who need to reuse features across projects and keep training and inference data aligned. Teams needing streaming or real-time feature sets should consider hybrid deployment.
Good to know, and what to watch
- Features can be reused across projects.
- Supports historical retrieval for training and online inference values.
- Lists a range of batch data sources.
- Supports deployment on cloud platforms or customer infrastructure.
- Offers a trial.
- JFrog ML SaaS supports batch feature sets only.
- Streaming and real-time sets require hybrid deployment.
Plans
Every plan and priceJFrog ML
Getting started
- Open the JFrog ML website
- Select the Start a Trial option
- Choose the SDK, CLI, or web interface for setup
- Connect a documented data source
- Define feature sets
- Deploy on JFrog's platform or customer infrastructure
- Serve models through an API, batch job, or Kafka application
Limits to check
JFrog ML SaaS supports batch feature sets; streaming and real-time feature sets require hybrid deployment. The JFrog ML plan has no listed price and directs buyers to contact sales.
The verdict
JFrog ML Feature Store is geared toward production ML teams that need consistent, reusable features. The SaaS feature-set limitation matters: streaming and real-time sets require hybrid deployment, and pricing requires a sales inquiry.
Questions people ask
Does JFrog ML have a free trial?
Yes. The product page offers a Start a Trial option.
Which platforms does JFrog ML support?
It supports API, Linux, self-hosted, and web access, with deployment on AWS, Google Cloud, Microsoft Azure, JFrog's platform, or customer infrastructure.
What data sources are supported?
Listed batch sources include Snowflake, BigQuery, MongoDB, Amazon S3 files, Redshift, MySQL, Postgres, ClickHouse, Vertica, and AWS Athena. Kafka and Kinesis are listed as streaming sources.
Can teams use both online and offline stores?
Yes. JFrog ML lists online store, offline store, and point-in-time joins.
How can models be served?
Models can run as live API endpoints, batch inference jobs, or Kafka stream applications.
What the maker publishes
- Data integrations
- Documented batch data sources include Snowflake, BigQuery, MongoDB, Amazon S3 files, Redshift, MySQL, Postgres, ClickHouse, Vertica, and AWS Athena. source
- Deployment
- JFrog ML supports AWS, Google Cloud, and Microsoft Azure, and offers deployment on JFrog's platform or on a customer's infrastructure. source
- Feature reuse
- Teams can define features once and reuse them across projects, with offline historical retrieval for training and online values for inference. source
- Feature Store
- The Feature Store centralizes development, sharing, and serving of ML features to keep training and inference data consistent. source
- Feature types
- Feature sets can be batch, streaming, or real-time, but JFrog ML SaaS supports batch feature sets only; streaming and real-time feature sets require hybrid deployment. source
- Feature workflow
- Users can define data sources and feature sets with the JFrog ML SDK or CLI, or create data sources in the JFrog ML UI. source
- Founded
- 2008 source
- Headquarters
- Sunnyvale, California, and Netanya, Israel source
- Intended users
- JFrog ML is presented for data scientists, ML engineers, and AI developers taking AI and ML services into production. source
- JFrog integration
- The Qwak integrations page lists JFrog Artifactory and Xray for model artifact management and scanning. source
- Model serving
- Models can be deployed as live API endpoints, batch inference jobs, or Kafka stream applications, with gradual deployment and A/B testing described in the product documentation. source
- Product scope
- JFrog ML provides tools to build, deploy, manage, and monitor AI workflows from GenAI and LLMs to classic machine learning. source
- Security compliance
- JFrog's Trust Center lists SOC 2 Type II and ISO 27001 among its certifications and reports. source
- Streaming integrations
- The data sources page identifies Kafka and Kinesis as streaming sources. source
- Trial
- The JFrog ML product page offers a Start a Trial option. source
Checked against the maker's pages on 8 Oct 2026.