Chalk

Paid#3 of 20 in Feature Store Software

Chalk: Advanced serving, versioning, and observability for production ML teams. Ranked #3 of 20 in Feature Store Software by our editors (9.0/10); pricing: Pricing on request; best for teams needing advanced serving and observability.

9.0/10★★★★☆Editor score

Visit Chalk

Chalk9.0 Visit Chalk

At a glance

• Editor score
  9.0 / 10
• Pricing
  Pricing on request
• Best for
  Teams needing advanced serving and observability
• Deployment model
  Both
• Facts checked
  21 Sep 2026

• Where it wins

  • Supports online and offline retrieval with point-in-time-correct datasets
  • Combines versioned catalogs, branching, and reusable feature definitions
  • Adds metrics, tracing, dependency graphs, and runtime dashboards

• Where it doesn't

  • Pricing requires a contact-sales conversation
  • Self-hosted deployments place infrastructure-policy responsibility with the customer
  • Its breadth may exceed the needs of teams seeking only a basic feature registry

Our verdict on Chalk

Chalk is a feature store and real-time machine learning platform for data scientists, ML engineers, and production AI teams. It lets teams define features once and reuse them across training, batch scoring, and online inference. The platform supports point-in-time-correct training datasets, low-latency online serving, offline queries for historical data, and both cloud and self-hosted deployment. Its Python and SQL feature pipelines make it suited to teams managing production machine learning workflows across development and serving.

Chalk’s strongest area is feature lifecycle depth. A versioned feature catalog combines metadata with feature branches for experimentation, while reusable definitions connect training, batch scoring, and serving workflows. Point-in-time-correct datasets support historical and training use cases, and the platform provides both online and offline stores. For teams operating real-time models, feature-level metrics, tracing, dependency graphs, and runtime dashboards add visibility beyond basic feature storage. These capabilities make Chalk a fit for organizations that need serving and observability in the same system.

Pricing requires contacting sales, so teams should expect a product and commercial evaluation rather than a self-serve plan comparison. Deployment is available through Chalk’s cloud or in a customer-controlled AWS, GCP, or Azure environment; self-hosted deployments place infrastructure-policy responsibility with the customer. Chalk is a strong choice for production AI teams that need advanced retrieval, dataset correctness, versioning, and operational visibility. Smaller teams seeking a simple feature registry, or buyers who prefer transparent plan-based pricing, should consider a narrower or more self-serve option.

Your product? Our editors can test Chalk in depth and publish a review, labelled as paid · from $149/mo →

Chalk pricing

Chalk fact sheet

Alternatives to Chalk

• Hopsworks Feature StoreA governed feature store spanning training, serving, monitoring, and deployment.9.4
• FeastA broad, flexible feature store for teams managing training data and online inference.9.3
• Databricks Feature StoreA broad, governed feature platform for teams already working in Databricks.8.9

See all Chalk alternatives →

Chalk vs the competition

• Chalk vs Hopsworks Feature Store
• Chalk vs Feast
• Chalk vs Databricks Feature Store
• Chalk vs Snowflake Feature Store
• CChalk vs Chronon
• FChalk vs Feathr

Compare Chalk with any tool side by side →

Featured on iTechGuides

Chalk is listed in our Feature Store Software directory. Add the badge to your site — it links back to this page.

<a href="https://www.itechguides.com/products/chalk/"><img src="https://www.itechguides.com/best/badge/chalk.svg" alt="Featured on iTechGuides" width="230" height="46"></a>

Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026

Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.

Last updated 7 Oct 2026 · How we research and update