Product
DataPilot uses real workspace and schema context to support engineering workflows around metadata, SQL, and result understanding. It is designed to work inside an operational product, not as a disconnected chatbot.
Schema-aware assistance
Schema context is injected before each AI request, including table names, data types, indexes, foreign keys, and existing comments. This removes hallucinated column names and keeps the workflow close to the catalog and workspace.
Documentation workflows
Generate publication-grade documentation from real schema. Review the delta and apply changes through ALTER TABLE and COMMENT ON statements directly from the workspace. Documentation stays connected to the actual database structure.
SQL support
Get context-aware optimization suggestions grounded in your schema, indexes, and key relationships. Click Fix with AI to get a corrected query using your real schema and runtime error details. No invented column names, no disconnected generic advice.
Why this page is not the homepage
AI assistance is one part of DataPilot's value. Teams also need scheduled delivery, private network execution, access boundaries, and operational visibility. Those needs do not go away just because AI can write SQL.
Does the AI know my actual table and column names?
Yes. Schema context is injected before AI requests, including table names, data types, keys, and existing comments.
Does AI documentation write to the database?
Yes. After review and approval, DataPilot applies comment updates through DDL statements.
What are AI credits?
Credits are the included AI usage budget on paid plans. Free customers can still use AI features with their own Anthropic API key.