Here’s a bad first draft of questions I could ask interview partners:

General Context Questions

Goal: Find out who I’m talking to.

What does your company do?
What do you do at your company?
- what’s the biggest challenge you’re facing in your day-to-day?
Tell me about your background. What did you specialise in? Where did you work before? What did you do there?

Data Questions

Goal: Find out about the problems they have.

Does your company have evolving datasets?
- background: we don’t care about static datasets, documentation for them is a solved problem.
Do they require non-standard analysis?
- background: for things like sales data, there’s a lot of standard tooling that can be used. no special documentation necessary.
  
  Research: Is Sales Data actually standardised?
What does your data stack look like? What infrastructure & tools do you use? Are you happy with them?
What does your data onboarding look like?
- How do new hires ...
  - ... learn about patterns & biases in the data?
  - .... learn how to interact with the data infrastructure?
Tell me about a time when your data evolved (aka patterns in data changed).
- How did you learn about it?
- What did you do about it?
- Are you happy with how that went?
  - Do you remember a time where that went worse / better? Tell me!

Pick thoughts on tool Idea

Goal: Present them that idea I have about bringing unit tests to data.

Idea: A small Python library with Assertions on statistics, similar to Great Expectations. Has support for Jupyter notebooks. Shows visual feedback on wether expectations are true, similar to Unit tests. Test results are collected, failing tests are notified via email.

some people say that unit tests can be useful both as verification + documentation. What would this look like, when applied to evolving data?
<tell more about idea>, <show mock-up>
1. Who do you think would this help?
2. Who wouldn’t this help?
3. How would this help you or your colleagues?

Outro

who should I talk to next?