jevsql

Small app.
Large searches.

Tell Jev what to find. Watch the matches arrive.

18 sample products
Ready to search.3 searches per day

Sample records 18

Connect your database

PostgreSQL, MySQL, MariaDB, MongoDB, libSQL, or an HTTPS SQLite file.

Upload a SQLite file

Use a read-only account on a public host. Selected data goes to Jev. Credentials stay in memory during the search.

PostgreSQL and MySQL try TLS first. If the server lacks TLS, the connection is unencrypted.

Scan settings

3 searches per IP address each day. 30 searches shared by all visitors. Limits reset at 00:00 UTC.

Each search checks up to 10,000 rows. A search can stop earlier at its token or time limit.

Keep this tab open. Results expire after 10 minutes. Changing data during a scan can affect pagination.

Record details

Use the API

Start a scan. Use the returned token to request progress and matches.

The public service has shared usage limits. Download the source to run your own instance.

How it works

  1. Choose a sample or connect your read-only database.
  2. Describe a match. Jev evaluates selected values in small, concurrent batches.
  3. Follow the scan. Change the threshold or stop when you have enough results.

Match scores are model estimates. The scan retains up to 1,000 highest-scoring records within a memory limit.

Large searches have real time and token costs. The movie shortcut uses a public CTU IMDb database.

View the movie dataset ↗