Glossary · AI & ML

Database Agent

A database agent is an AI system that can autonomously perform multi-step work against a database–planning queries, running them, inspecting results, and refining its approach–rather than answering one prompt at a time.

A database agent goes beyond single-shot question answering. Asked "why did signups drop last week?", an agent might first query daily signup counts, notice the drop started Wednesday, then query by acquisition channel, find one channel collapsed, and report the finding–a chain of queries a human analyst would run. This matters because real analytical questions rarely have one-query answers. Agents handle the iteration, which is where most of an analyst's time goes. The trade-off is trust: agents need guardrails like read-only access and query validation so autonomy never puts data at risk.

Fig — every answer ships with the tables, rows and SQL behind it.