Scenario cases¶
Scenario cases help users understand how DataFoundry handles realistic data-agent tasks across schema inspection, read-only SQL, trace evidence, and final operating decisions.
What makes a good case¶
A strong case should be concrete enough to reproduce and polished enough to show externally:
- A recognizable business or operational scenario.
- A small demo-safe dataset or schema outline.
- A clear analysis path, including the questions a user would ask.
- Expected findings that can be validated from the data.
- Notes about governance, privacy, access control, or production boundaries.
The DTC growth operating review is the current reference case. It combines orders, ad spend, products, support tickets, and targets into one traceable operating review.
How to propose one¶
Open a GitHub issue with:
- Case title and target audience.
- Business question or workflow goal.
- Tables, files, APIs, or external systems involved.
- Expected analysis steps and representative outputs.
- What must be anonymized, mocked, or excluded.
Cases should not require private data access. If the real scenario is sensitive, describe the workflow shape and provide a sanitized dataset or synthetic equivalent.