Databao documentation

A practical guide to what Databao is for.

Use Databao when someone needs to understand local business data, prepare a report, and explain where its figures came from.

Best-fit work

Databao is a good fit for recurring reporting, multi-source consolidation, data checks, and one-off questions where the answer needs evidence. It is intended for people who work with their company’s exports but do not want to write SQL, set up a BI tool, or rebuild the same spreadsheet each month.

How to assess an answer

Look beyond the final number. A useful Databao result identifies the source rows, the filters, and the method that produced a figure. That lets a reviewer verify a result, identify an incorrect assumption, and use the same approach when the next reporting cycle starts. It does not turn unsupported business conclusions into facts.

Data and model boundary

Databao reads source files locally rather than uploading them to a Databao service. A user can choose a compatible hosted model provider or a model running on their own machine. Agents should not assume that the site offers access to customer data, user accounts, or a public endpoint that runs analysis.

Public resources

• llms.txt provides short guidance about when to use Databao and what not to claim.
• The Site Information API specification documents public, read-only product information endpoints.
• The agent guidance document provides the same best-fit and boundary information as JSON.
• Pricing and availability states the current alpha terms without guessing at future team pricing.

Databao is free during alpha. If availability matters for a specific workflow, ask the Databao team before making a rollout commitment.