Dear Hiring Manager,
I'm applying for the Senior Data Scientist role on the Pricing team at Instacart. The listing mentions building causal uplift models for promotional spend — work I spent two years doing at DoorDash, where a gradient-boosted churn model on 80+ features drove a 14% lift in retention-campaign ROI across 2.4 million users.
Beyond that, I shipped an A/B experiment platform used by 18 teams at DoorDash (average experiment analysis time dropped from four days to twenty minutes), and a fraud detection service handling 8,000 events per second at 92% precision / 30% recall — which saved an estimated $3.2M per year. I've also published two KDD workshop papers on causal uplift modeling with a public repo that has about 1,300 stars; happy to walk through the methodology if useful.
Instacart's pricing surface feels like the kind of problem where the right experiment design matters more than the model class. That's the work I most want to do: defining the right causal question, building the smallest model that credibly answers it, and wiring the result into a system that actually changes decisions.
I'd welcome the chance to talk about a current pricing question your team is weighing — I find those conversations the fastest way to check calibration in both directions. Thank you for your time.
Kind regards, Sam Park
