🚚 Deliver

Publishing is a governed act, not an export

The last step of most data work is someone emailing a spreadsheet, at which point every guarantee the pipeline made stops applying. DataLens makes delivery part of the platform: a data product with an owner and a contract, or a share whose access you can see and revoke.

Two grain sizes

Data products

A curated, owned, documented dataset published for other teams to consume — with a stated contract, a refresh cadence and a named owner. This is the data-mesh shape: the team that knows the data publishes it.

Publish and share

For the case that does not need a product: an access-controlled link, an export, or an API endpoint serving the current version of a dataset.

What travels with a published product

The contract

What consumers are promised: the columns, their types, and what will not change without notice.

The lineage

Which sources it came from and what was done to them, so a consumer can judge whether it answers their question.

The quality record

The rules the data passed, and when they were last evaluated.

The owner

A person to ask, recorded with the product rather than known informally by four people.

Access control

Who can read it, granted explicitly and revocable, with the grants visible in one place.

A version

Products are versioned, so a consumer built against last quarter’s shape is not broken silently.

Delivery questions

What is a data product, concretely?

A dataset published with a named owner, a stated contract, its lineage, its quality record and explicit access grants. The difference from an export is that all of that travels with it and stays current.

Can other systems read a published dataset?

Yes. Published assets are served over an API with issued keys, so a downstream system reads the governed version rather than a copy of a copy.

Can I revoke access after sharing?

Yes. Grants are explicit and listed, and revoking one takes effect for the next request rather than requiring you to find every copy.

See it on your own data

DataLens is in private beta. Bring a file, a database or an API and work through the whole lifecycle in one sitting.

Request beta access