On 2026-03-23 16:46:42, Peter Karavias via Wellfound Match wrote:
Hey Robert!
Thanks for applying.
We’ve been seeing a high volume of AI-generated and proxy profiles, so before moving forward we ask all candidates to answer a few quick questions.
Please keep answers concise but specific.
1. Describe one production system you personally built or owned that involved messy or unstructured data (e.g., logs, documents, exports, etc).
What was the raw input?
How did you transform/clean it?
What did the final structured output look like?
What broke or was hardest?
2. Share one GitHub repo OR code snippet that best represents your backend/data work.
Tell us exactly what part you personally wrote
If it’s a team repo, specify your contribution
3. You receive 10,000 messy CSV files from different distributors, all with different schemas.
How would you design a pipeline to normalize them into one schema?
What tools + steps would you use?
(No need for perfection or exhaustive answers— we’re looking for how you think.)
Thank You,
Tritone Team
View Conversation