Database · Connector
How do we connect PostgreSQL or MySQL to AI agents on our own infrastructure?
Katonic connects to a PostgreSQL or MySQL database, runs the queries you write and indexes each row they return as a document in the Knowledge Engine, on infrastructure you control. It suits knowledge kept in tables, such as a wiki, a set of questions and answers or a content system.
What it reads.
From your database.
- 01One document for each row that a query returns
- 02The columns you choose, joined into the text of the document
- 03An HTML column, kept as HTML so that it is split at its headings
This connector is for knowledge kept in tables. It indexes what your queries return; it is not how an agent runs a new query against the database.
Through: PostgreSQL and MySQL
Where the content goes
Into a knowledge store you run.Content is pulled into the Knowledge Engine on infrastructure you own or control, and indexed there. Agents answer from it with a model you serve. The documents, the index and the questions are not sent to a public AI service.
How it signs in
A host, port, database name, username and password.
The set-up dialog asks for
- Database type
- Host
- Port
- Database name
- Username
- Password
- SQL queries
Permissions
Access is set on the source in Katonic: everyone given access to the source can retrieve its documents. The permissions Database holds for each item are not read.
There are no permissions for each row.
Sync
Full, by running the query again, or incremental by an updated-at column.
You set how often the source is read again, in minutes, hours, days or weeks.
Limits
- Each query returns an id, a title and a content column.
- A database on a private network address needs an operator to add it to an allowlist first.
Use case pages that name Database. More are in the gallery.
- Data Analyst Agent →
- KYC/AML Compliance Agent →
- Fraud Detection Agent →
- Fraud Detection Agent →
- Benefits Processing Agent →
- Compliance & Reporting Agent →
- Policy Analytics Agent →
- Infrastructure Management Agent →
- Network Optimization Agent →
- Churn Prediction Agent →
- Fraud Detection Agent →
- Billing & Revenue Agent →
Which databases does it read?
PostgreSQL and MySQL.
Does it index whole tables?
It indexes what your queries return, one document for each row. You write the queries.
Where are the rows indexed?
In the Knowledge Engine, on infrastructure you own or control. They are not sent to a public AI service.