Like ChatGPT, but for your database.
Correct answers from
your database.
On the first try.
CheckMyData reads your schema and your codebase, so it knows your cents-vs-dollars, soft-deletes, and enums. Ask in plain English; get SQL and answers you can trust.
No credit card. Deploy in minutes. Self-host or use hosted.
The context engine
Why the answers are actually correct
Data is abundant. The context needed to query it correctly is scarce. CheckMyData assembles that context before it writes a single line of SQL.
- 01
Your schema
Tables, columns, types, and relationships — read directly from your database.
- 02
Your codebase
Business logic from your repository: what “active”, “amount”, or “deleted” really mean.
- 03
Your rules
Conventions and guardrails you set, applied to every query it writes.
- 04
Its memory
Patterns it has confirmed on this connection over time, reused on the next question.
Validated, dialect-aware SQL
The query is checked against your real schema and run for you — and if it fails, it self-heals and tries again. You get an answer you can trust, with the SQL shown so you can verify it.
vs. a plain SQL editor
- Has no idea what your columns mean — you supply all the context.
- You hand-write and debug every query yourself.
- CheckMyData knows your schema and code, then writes the SQL for you.
vs. a generic chatbot
- Guesses your schema and invents table and column names.
- Can't run the query or check whether the result is right.
- CheckMyData grounds every answer in your real schema, then runs and validates it.
Everything you need to actually
understand your data
Context-aware answers, self-healing queries, and explanations you can trust — backed by readable SQL and exportable charts.
Codebase-aware context
It indexes your repository, so answers reflect your real business logic — soft-deletes, enums, money-in-cents — not just raw column names. That context is the difference between a correct answer and a confident wrong one.
Self-healing queries
When a query errors, it reads the failure, pulls in the missing context, and repairs itself — so you get a working answer instead of a stack trace.
Institutional memory
It learns the patterns of each connection over time, confirms what's right, and lets go of stale assumptions — so the more you use it, the sharper it gets.
Natural-language queries
Ask in plain English and get back the SQL, the result, and an explanation — so anyone on the team can answer their own data questions.
Multi-database support
Connect PostgreSQL, MySQL, ClickHouse, and MongoDB — directly or over an SSH tunnel — and query them all the same way.
Automatic visualization
Charts, tables, and exports are generated for you, so you see the shape of the answer without building a dashboard.
Self-hostable
Deploy anywhere with Docker and keep full control of your environment — no vendor lock-in, no data leaving your infrastructure.
Privacy-first by design
No advertising or analytics trackers, read-only by default, and errors reported with secrets scrubbed twice — the source is open, so you can verify exactly how your data is handled.
What can you do with it?
Ask questions in everyday language and get back the rigor of a data scientist: the query, the result, and the reasoning behind it.
- Find why revenue dropped last week
- Analyze user behavior without writing SQL
- Explore data across multiple databases
- Debug backend logic with real data
- Build internal dashboards in minutes
you: Why did revenue drop last week?
ai: Analyzing orders table...
SELECT date, SUM(amount) FROM orders
WHERE date >= '2026-03-17'
GROUP BY date ORDER BY date
ai: Revenue dropped 23% on March 20.
Root cause: payment gateway timeout affected 142 orders.
+ Chart generated · Exportable
From zero to insight in 30 seconds
Connect
Plug in your database. Securely.
Ask
Type your question in plain English.
Get answers
Queries, charts, and insights, instantly.
Built in public, fully transparent
The entire source is on GitHub under the MIT license.
Inspect exactly how your data is processed, contribute improvements, or self-host on your own infrastructure.