From the makers of ReinDB
ReinQ: ask your own data in plain English Working prototype
ReinQ is a working prototype, built alongside ReinDB: try it on demo data now, or ask us to connect your own database. Ask a business question such as “monthly sales this year for each product” and get the answer, the chart and the SQL behind it, straight from your own PostgreSQL, MySQL, SQL Server or Oracle database, a warehouse such as Snowflake, BigQuery, Redshift or Databricks, a data lake through Trino, Starburst or Athena, or a spreadsheet you upload. Real schemas are messy, with cryptic table names, backup copies and three different customer tables, so ReinQ asks when it is not sure, remembers your answer, and learns your schema the more questions you ask.
Learns your schema as you ask
It asks which table really holds your customers or what “sales” means to you, once, and every later question uses that answer.
Narrow it down in plain words
Follow up with “only the West”, “just last quarter” or “top 20” until the result is the one you need.
Shows its work, ready to share
Every answer comes with its SQL and checks, and turns into a formatted PDF report in one click. Excel and CSV export come next.
Finds the joins nobody wrote down
No foreign keys? ReinQ infers how tables link and flags weak matches, like a backup copy that looks like the real thing.
Checked before it runs
Read-only, with a time limit and a 1,000-row cap. The plan’s cost is checked first, and a full scan of a big table is refused with a suggestion. Any query still uses the server, so point ReinQ at a read replica when you can.
Its own login, kept apart
ReinQ reads rows through its own read-only login. ReinDB still never reads your table rows.
Planned, beyond the preview
Your databases, warehouses and lakes
PostgreSQL, MySQL, SQL Server and Oracle, on your servers or on RDS, Aurora, Azure and Cloud SQL; Snowflake, BigQuery, Redshift and Databricks; Trino, Starburst and Athena over a lake; or a CSV or Excel file you upload. Read-only logins only, and on a warehouse the scan size and cost are checked first.
Schema discovery
Reads the catalog and comments, decodes legacy abbreviations (cust, ord, amt, dt, mstr), infers joins, labels status codes and skips junk tables.
A team glossary
Answers people give are kept for each database and used for everyone who asks about it; your DBA approves, edits or rejects them on one review screen.
Your AI model, or none
The same model choice as ReinDB, including a local model. Only the question and schema names go to it, never result rows.
Saved and scheduled
Save a question, share it, and email the formatted report on a schedule, such as monthly sales on the 1st.
Audit and sensitive data
Every question, its SQL and row count in an audit log. Emails, phone numbers, tax IDs and card numbers are masked by default, found by column name or by what a column holds; when staff choose to see them, the log records it.
ReinQ is a working prototype; planned features and their timing may change. Want to try it on your own data? Email inquiries@bandlei.com.
Try ReinQ on your own schema
Tell us about your messiest database. We’ll set up an evaluation on a read-only login.