Data lake and analytics · AutessaDB
Data agents
An Autessa agent can answer data questions over the AutessaDB data lake. Create agent tools on a table makes three tools bound to the tables you pick, describe the tables, run a query, and get a query's results, and an agent with them looks up what the tables contain, writes SQL, prefers your verified queries, filters on partition columns and pages through long results.
-- the problem
Data questions wait in the analytics queue
"How many tickets did we close last month, by team?" is a five-minute query that can wait days for an analyst. A data agent answers it in conversation, from the same tables, with the same measures and the same permissions as the person asking.
What you get
- Create agent tools on a table's detail page, or create-agent-tools in the CLI, makes all three tools at once.
- Describe returns SQL names, columns, partition columns, row counts, measures, joins and verified queries.
- Query runs one read-only SELECT as a job; get result pages through the rows.
- Agents prefer a verified query when a question matches one, so recurring answers stay consistent.
- Agents filter on partition columns to keep scans small, and page through long results.
- Agents never receive download links; they read results through the tools.
- An agent sees only what the person or key it acts as could see: sharing, row rules, hidden columns and masks all apply.
-- how it works
How does data agents work in AutessaDB?
Describe the data
Fill in the semantic layer, or draft it with Suggest with AI.
Create the tools
Click Create agent tools and pick the tables the agent should use.
Attach them
Add the three tools to an agent.
Ask away
The agent describes, queries and reads results, then answers in plain language.
-- every option
The three tools
| option | what it does |
|---|---|
| Describe tables | What the tables hold: SQL names, columns, partitions, measures, joins and verified queries. |
| Run a query | One read-only SQL query over the bound tables, as a job. |
| Get results | A page of a finished query's rows. |
-- example
Full reference in the AutessaDB CLI docs and the API reference.
-- permissions
Who can do what?
| access | can |
|---|---|
| The agent's identity | Needs READ on every table it queries; sees what that identity may see. |
| Builders | Create the tools and attach them to agents. |
◆ AI assist
Where AI helps
The agent decides when to describe, when to query and when to page, and leans on the semantic layer and verified queries so its SQL matches how your team measures things.
AI in AutessaDB works inside the same access rules as the person using it, and schema changes always wait for a person to apply them.
-- faq
Questions about data agents
Can an AI agent answer questions about my data lake?
Yes. Give it the three generated tools and it describes the tables, writes SQL, runs it and reads the result.
Can a data agent see data the person asking cannot?
No. The agent acts as a person or key and sees only what that identity could see.
How do agents keep answers consistent?
They read your measures and prefer verified queries when a question matches one.
100% customizable, code and all
Make AutessaDB fully your own.
Use it out of the box, then change anything. Ask Autobot, Autessa’s AI builder, for a new field, screen, report, workflow or agent and it builds it into your copy of the app. Or open the code and edit it yourself. Nothing is locked: the app, its data and its code are yours.
Start free and make it yours