Senior Machine Learning Engineer, Personalization & Decision Science

Responsibilities: Build Cooklist's recommendation stack: design and productionize multi-stage systems for candidate generation, retrieval, filtering, scoring, ranking, reranking, and final selection across recipes, products, meals, collections, substitutions, and replenishment items. Model household preferences: learn individual and shared household affinities from explicit preferences, order history, recipe interactions, cart edits, substitutions, removals, repeat purchases, and conversational feedback. Distinguish taste from circumstance: identify when an interaction represents a durable preference versus a temporary need, guest purchase, holiday, party, work lunch, children's food, one-time recipe, or another contextual exception. Optimize complete baskets: formulate meal and grocery planning as constrained, multi-objective problems balancing household value with budget, servings, nutrition, allergens, convenience, variety, ingredient reuse, package sizes, pantry carryover, leftovers, and availability. Predict replenishment and substitutions: estimate recurring needs from purchase timing, quantity, household size, seasonality, cadence, and uncertainty; rank replacements using functional equivalence, recipe context, diet, brand, price, package size, inventory, and predicted acceptance. Handle cold start and catalog variation: personalize for new households, members, recipes, products, and sparse users while connecting canonical foods to store-specific SKUs and adapting to assortment, price, promotions, out-of-stocks, regional availability, and catalog drift. Evaluate at the basket level: go beyond item relevance to measure constraint violations, diversity, calibration, coherence, waste, ingredient reuse, acceptance, cart edit distance, substitution acceptance, order behavior, repeat usage, and retention through offline replay and online experiments. Take models to production: own pipelines, batch and low-latency inference, caching, versioning, monitoring, drift detection, fallbacks, and rollback; expose candidates, scores, confidence, constraints, reason codes, and alternative plans through services Cooklist's applications and AI agents can safely consume.

Qualifications: You have typically 5+ years of experience in machine learning engineering, applied science, recommender systems, operations research, or a related quantitative field. You have shipped a meaningful recommendation, ranking, forecasting, personalization, or optimization system into production and owned its performance after launch. You have deep expertise in recommender systems, retrieval, ranking, embeddings, or user modeling and meaningful working knowledge of forecasting, mathematical optimization, or other decision-science methods. You understand modern recommendation methods such as collaborative filtering, content-based models, two-tower retrieval, sequence models, learning-to-rank, reranking, calibration, exploration, and contextual bandits - and know when a simpler approach is better. You are highly proficient in Python and SQL and write maintainable, rigorously tested production code - not only research notebooks. You understand implicit-feedback problems including exposure bias, negative sampling, delayed outcomes, sparse labels, popularity effects, feedback loops, and the difference between 'not selected' and 'disliked.'You understand experimentation and causal reasoning well enough to distinguish genuine model improvement from noise, leakage, bias, or misleading offline movement. You can operate across the production ML lifecycle, including feature freshness, training-serving consistency, latency, caching, monitoring, data and model drift, graceful degradation, and retraining. You communicate clearly with engineers, product leaders, culinary stakeholders, and founders, and you care about what makes a meal complete, a week varied, a substitution valid, and a shopping cart believable.

Bonus Points: Experience in grocery, meal kits, food delivery, retail, e-commerce, marketplaces, subscriptions, or another constrained recommendation environment. Experience with consumption or repeat-purchase forecasting, survival analysis, mathematical programming, constraint optimization, or real-time recommendation systems. A master's degree or PhD in computer science, operations research, statistics, applied mathematics, industrial engineering, economics, or a related field. Our Stack Application & interfaces: Python/Django backend; Java Script/React Native frontend; GraphQL and real-time streaming over Web Sockets Data domain: recipes, ingredients, products, retailer catalogs, transactions, carts, prices, promotions, inventory, nutrition, preferences, and pantry signals Modeling: Python-first recommendation, ranking, forecasting, experimentation, and optimization workflows using the right mix of modern ML, classical statistics, and mathematical programming Serving & evaluation: batch and real-time model services integrated with Cooklist's Django backend; offline replay, synthetic household scenarios, basket-level regression suites, online experiments, and production monitoring

What We Offer: Competitive compensation + meaningful equity Austin, TX based with WFH flexibility Work directly with founders and an elite, tight-knit team Ship products and experiences that materially improve the lives of millionsA high-intensity, high-ownership environment designed for builders