Features

Everything between the source and the answer

DataLens is organised around the lifecycle a piece of data actually travels, not the org chart of the team that owns it. Each stage below is a working surface in the product.

Capabilities that cut across every stage

Ari, the data assistant

Reachable from every screen. It has the profile of the active dataset in front of it, so its answers are about your columns rather than about data in general.

Workspaces

A workspace holds the datasets, models, flows and releases for one piece of work, and switching between them swaps the whole context at once.

Versioned releases

A release seals models, mappings, prompts, golden datasets and flows together, then promotes them through environments as a unit.

AI operations

Cost, token use, latency, failure taxonomy, cache hit rate and estimate-versus-actual for every AI step the platform runs for you.

Prompt library

Prompts are versioned artefacts bound to flows, not strings pasted into a node and forgotten.

Document generation

Turn a governed dataset and its lineage into a document — a data dictionary, a quality report, a handover pack.

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