We're Bespoke AI, a 3-person Bay Area startup building AI coaching software for restaurant servers. We're launching our first paid pilot with a Bay Area fast-casual chain (6–7 locations) and need a statistician or data scientist to design and analyze a causal study of our intervention. The project; We want to know whether our coaching software causally lifts two metrics: average check size and items per transaction. Our design partner has offered 24 months of historical POS data and is willing to participate in study design. Two phases, ~30–40 hours total Phase 1 (next 2 weeks): Recommend study design (DiD vs. synthetic control vs. other), audit the historical POS data, compute power, and pre-register an analysis plan before the pilot launches. Phase 2 (after pilot): Run the pre-registered analysis and deliver a causal writeup with effect estimates, confidence intervals, and limitations. We're looking for someone with: Causal inference experience (DiD, synthetic control, matched controls, or RD) Quasi-experimental design at small N Power analysis / sample size calculation Python or R Ability to communicate clearly with a non-technical CEO Budget: Up to $5,000 fixed or hourly with cap. Structured as two milestones. To apply: Briefly describe one project where you measured the causal impact of an intervention on a business metric — method used and its main limitation. Two paragraphs is enough. We need a scoping call before Friday May 15. Strong applicants hear from us within 24 hours.