Snowflake · Connector
How do we make our Snowflake catalogue searchable by AI agents on our own infrastructure?
Katonic connects to Snowflake and indexes its metadata: tables, views, columns, comments, stored procedures and functions. An agent can then answer which table holds what, with a model you serve. The connector indexes the catalogue, not the rows in your tables.
What it reads.
From Snowflake.
- 01Table and view metadata: columns, types and comments
- 02Stored procedures and functions
- 03Table comments and column descriptions
- 04Listings of stage files, where you turn that on
The connector indexes Snowflake's metadata and documentation, not the data in your tables. It lets an agent answer which table holds customer revenue, or what a column means.
Through: Snowflake, by metadata queries
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
In the set-up dialog you give a username and password. The connector also supports key-pair sign-in.
The set-up dialog asks for
- Account identifier
- Username
- Password
- Warehouse
- Databases
Permissions
Access is set on the source in Katonic: everyone given access to the source can retrieve its documents.
Sync
A full refresh of the metadata on the schedule you set.
You set how often the source is read again, in minutes, hours, days or weeks.
Use case pages that name Snowflake. More are in the gallery.
Does it copy our data out of Snowflake?
No. It indexes metadata and documentation: table and column names, types, comments, procedures and functions. It does not index the rows.
Can we limit it to some databases?
Yes. You list the databases to read.
Where is the metadata indexed?
In the Knowledge Engine, on infrastructure you own or control. It is not sent to a public AI service.