SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Lab37 Robotics is building robotic solutions for direct-to-customer food production. This Senior Data Scientist role owns the forecasting and optimization models that drive operational decisions: prep quantities, robot refill schedules, and labor scheduling against demand.
The core challenge is decision-making under uncertainty—committing to prep before demand is known, then scheduling constrained, perishable production against those commitments. This is a stochastic optimization problem with a probabilistic-forecasting front-end.
You will formulate and solve the core planning problem as optimization under uncertainty, coupling prep, refill, and labor decisions through shared, perishable, capacity-constrained production. You'll define objectives over real tradeoffs (waste, stockouts, labor, quality, timing) under hard perishability constraints, and derive service levels rather than hand-tuning them. You'll match method to problem across closed-form, heuristic, and exact approaches (LP/MIP/CP), knowing when decomposition, duality, or a full solver is warranted.
You'll develop probabilistic demand forecasts for high- and low-volume components, produce calibrated forecasts, and correct censored demand where stockouts hide true demand. You'll solve cold-start forecasting for new locations with little or no history.
You'll build optimizers that continuously re-plan and that operators can read, trust, and override, with overrides captured as signal. You'll turn today's heuristics into constraint-aware optimizers and prove gains with walk-forward backtests that price tradeoffs in operational and dollar terms.
Cross-functional collaboration is essential: you'll partner with operations, product, and engineering to translate data into crisp requirements, and with data engineering on model inputs. You'll communicate complex concepts clearly to technical and non-technical stakeholders, with explicit assumptions and tradeoffs.
Required: Advanced degree (PhD or MS) in Operations Research, Industrial/Systems Engineering, Statistics, Computer Science, or related quantitative field. 3+ years in operations research, optimization, forecasting, applied modeling, or equivalent depth from graduate research. Demonstrated depth in optimization under uncertainty: formulated and solved stochastic, mixed-integer, scheduling, or inventory problems on real, messy data. Duality/shadow prices, decomposition, and solver tuning (Gurobi, CPLEX, CP-SAT/OR-Tools) expected. Probabilistic forecasting with decision-focused mindset: calibration, quantile accuracy, censoring correction, per-location bias, cold-start. Strong Python and SQL; production-ready code; comfort with an optimizer/solver. Systems thinking and commitment to interpretability—plans and forecasts operators can trust, debug, and override, not black boxes.
Desirable: Stochastic programming, robust optimization, or scheduling with sequence-dependent setups. Real-time operational or telemetry data in production. Food service, on-demand delivery, or just-in-time manufacturing experience. Shipping models behind an API with engineering.
