Data Discovery

The Data Discovery feature in Zerve eliminates the manual effort of mapping out new datasets. By connecting directly to your organization’s assets, it acts as an intelligent layer that inventories your data, understands relationships, and suggests analytical pathways, saving data professionals hours of exploratory work.

To begin, ensure your database should be connected in the Organization Assets. Once the connection is established, you can trigger Data Discovery to scan the schema. You can find documentation about this here

Method 1: Automated Data Discovery

Choose this for a comprehensive overview of a new or unfamiliar database. Zerve will perform an exhaustive crawl of the entire schema, identifying every table and mapping out the inherent architecture without requiring any manual input.

Method 2: Prompt-Guided Discovery

Choose this to narrow your focus to specific tables or business goals. By providing a natural language prompt, you can direct the agent to prioritize certain areas of the database.

The Discovery Output

Once the process is complete, Zerve provides a rich, structured profile of your data. The discovery output includes:

• Table Descriptions: Concise summaries explaining the purpose and content of each table, translating cryptic column names into clear business context.

• Relationship Mapping (Joining Keys): Automatically identified Primary and Foreign keys. This highlights how tables relate to one another, making it easy to build complex joins for your analysis.

• Data Types & Schema: A clean breakdown of types and formats within each column to ensure your queries are syntactically correct from the start.

• Suggested Use Cases: A list of specific "Questions that can be answered." This transforms the raw schema into a roadmap of potential insights (e.g., "Customer purchasing patterns and segmentation").

Following video shows how Data Discovery feature can be used: