Ask your database in plain English.
Joinery turns a question into SQL, runs it against your database or warehouse, and shows you the query next to the answer — so you can check the reasoning instead of trusting it.
Not generally available yet. No card, no signup — just an email.
Which customers spent the most last quarter?
SELECT c.name, SUM(o.total) AS spend FROM orders o JOIN customers c ON c.id = o.customer_id WHERE o.placed_at >= date_trunc('quarter', now()) - interval '3 months' GROUP BY c.name ORDER BY spend DESC LIMIT 10;
| Customer | Spend |
|---|---|
| Northwind Traders | $412,900 |
| Contoso Ltd | $388,140 |
| Fabrikam Inc | $301,775 |
Illustration of the intended interaction. Sample data, not a live query.
Exploring data shouldn't depend on remembering the schema
Most questions about a database are easy to ask and tedious to write. Joinery closes that gap without hiding what it did.
Plain English in, SQL out
Ask the way you would ask a colleague. Joinery reads your schema — tables, columns, keys, types — and writes the query that answers the question.
The query is always shown
You see the SQL before and after it runs. Copy it, edit it, paste it into your own tooling. Nothing is a black box you have to take on faith.
Exploratory, not one-shot
Follow up in the same thread. Narrow a filter, change a window, group by something else — without rebuilding the query from scratch each time.
Reads, by default
Connect with a read-only role and Joinery can explore but not alter. Your database's own permissions stay the boundary.
Built for people who already work with data
Not a replacement for knowing SQL — a way to spend less time on the boilerplate around the interesting question.
Developers
- Check data without leaving what you were doing
- Draft a query against an unfamiliar schema
- Sanity-check what a migration actually changed
Analysts
- Answer ad-hoc questions without a ticket queue
- Run exploratory analysis across a warehouse
- Keep the SQL for the questions worth repeating
Data & platform teams
- Shorten the ad-hoc request backlog
- Hand out read-only access with real boundaries
- Let people self-serve without new BI seats
In development
Joinery is being built and is not generally available. Nothing is for sale yet and there is no pricing page — when that changes it will be published on dfacto.ai/pricing before anything can be bought.
If the problem above is one you have, say so. Early access goes out in small batches, and the people who describe their setup get in first.