Operational questions often start in MySQL. An order is waiting, an account changed, or an inventory count no longer matches the number in front of you. The useful answer lives in the database serving the application, not always in a warehouse built to receive it later.
GenViz could already reach Postgres, MongoDB, SQL Server, BigQuery, Db2, and AS/400. MySQL, the most widely deployed operational database, was the obvious gap in that list. You can now connect it and ask those questions in plain language.
Add the connection
Choose MySQL when you add a connection, then paste a mysql:// URI. The URI
includes the user, password, host, port, and database. The default port is 3306, and GenViz stores the credentials encrypted.
Once connected, you can ask about the records and relationships in the database
without writing SQL. GenViz automatically gives the model MySQL-specific
guidance for LIMIT and OFFSET pagination, information_schema introspection,
and backtick identifier quoting. Those details stay behind the question you
asked.
Run it where the database lives
MySQL connections support both GenViz runtimes from day one. As with Postgres, you can run a query locally from the desktop app or on a connector worker inside your network.
The connector matters when your database sits on a network your laptop cannot reach. The query runs from the worker that can reach it, while you keep asking questions from the app.
Keep every question read-only
MySQL follows the same read-only discipline as every other SQL source. A
question becomes one SELECT statement, or one WITH...SELECT statement.
Anything shaped like a write or containing multiple statements is refused
before it reaches the driver.
Results are bounded as well. You get enough rows to answer the question without turning a conversation into an unbounded pull from an operational system.
Turn an answer into a shared table
When an answer needs to travel beyond the conversation, pull it as a dataset. GenViz streams the result to Parquet under the hood and gives you a table you can share with your team.
MySQL extracts get the same version history and compare features as extracts from other sources. You can refresh the original question later, keep the table you had, and inspect what changed between the two results.
MySQL is available now in the connection type picker in the app. Setup details are also in the connections reference at /solutions/connections.