Business Insights AI Agent using CDGC Metadata Discovery > Introduction to Business Insights AI Agent using CDGC Metadata Discovery recipe > Business Insights AI Agent using CDGC Metadata Discovery recipe contents
  

Business Insights AI Agent using CDGC Metadata Discovery recipe contents

The Business Insights AI Agent using CDGC Metadata Discovery recipe contains a model connection, Python tools for metadata retrieval and query execution on Snowflake and Databricks platforms, and an agent flow with inline agents.
The following table lists the assets that the Business Insights AI Agent using CDGC Metadata Discovery recipe contains:
Asset Name
Asset Type
Description
Azure OpenAI LLM
Model connection
Model that connects to the Microsoft Azure OpenAI LLM.
Get_Business_Glossaries
Python Tool connection
Tool that retrieves CDGC business glossaries related to data elements within a specified catalog source.
Get_Tables_for_Business_Glossaries
Python Tool connection
Tool that takes business glossaries and a catalog source as the input, and returns the related tables with details, such as identity, name, description, database, and schema.
Get_Tables_Details_with_Columns
Python Tool connection
Tool that takes table identifiers with partial metadata as input and retrieves the related column IDs for each table. The tool calls a REST API to fetch complete column metadata, including the name, description, data type, and nullable status. The tool then constructs a detailed structure of the table, including its name, description, schema, database, and columns.
Get_Tables_Relationships
Python Tool connection
Tool that takes tables with partial metadata as input to retrieve the related foreign key IDs. The tool then uses a REST API to fetch detailed relationships between tables stored in CDGC and constructs an array of JSON structures with the table details and relationships.
Snowflake_Execute_SQL_Statement
Python Tool connection
Tool that executes SQL statements in Snowflake and returns responses. Input parameters include the SQL statement, schema, and database.
Databricks_Execute_SQL_Statement
Python Tool connection
Tool that executes SQL statements in Databricks and retrieves responses. Input parameters include the SQL statement, schema, and database.
AI Agent for Business Insights
Agent flow
Agent that extracts relevant metadata from CDGC based on a user's business question. Generates and executes SQL queries on the connected database, and returns both results and summarized responses.