The person with the question does the analysis
Self-service usually fails one of two ways: the data is governed but unusable, or usable but ungoverned. The gap is that governance and analysis live in different tools, so one of them always wins.
How DataLens approaches it
Governed data, directly usable
A published data product carries its contract, owner and lineage. Using it does not mean leaving the governed path.
The assistant knows the data
Ari has the profile of the dataset in front of it, so a question about your columns gets an answer about your columns.
Answers show their work
A number arrives with the aggregation that produced it, so it can be checked rather than trusted or dismissed.
Lineage travels with the chart
When a chart looks wrong, what was done to the data is one click away rather than in another team’s tool.
Sharing stays governed
A shared result is an access-controlled link you can revoke, not a spreadsheet in an inbox.
The stages this leans on
The rest of the lifecycle is there when you need it.
Questions
Do business users need training?
The assistant is the interface for most questions, so the floor is low. The people building the data products underneath do need to understand the modelling stage.
How do we stop people producing contradictory numbers?
The semantic model. Defining “active customer” once means every chart and product built on it agrees, which is the actual cause of contradictory numbers.
Can analysts still write SQL?
Yes. The assisted path is the default, not the only one.
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