Machine Learning Engineer

company descriptionthe analog company is redefining loyalty for hospitality — designed for cafes, bakeries, and restaurants that care about brand, experience, and community.we believe loyalty should feel like recognition, status, and belonging, not manipulation or promotions. we're designing an upscale, mobile-first loyalty experience that hospitality groups and culturally relevant brands would actually be proud to use.role descriptionwe're looking for a founding machine learning engineer to build analog's intelligence layer from the ground up, working directly with our CTO and founding team. this is a hands-on, high-ownership role for someone who wants to define what recognition looks like in code.at analog, ML is not a feature — it's the engine behind every guest moment. the signals we learn from a guest (how they visit, what they engage with, who they bring) are what turn first-time guests into regulars. you will own the models, the pipelines, and the judgment behind them.what you'll dodesign and ship the recognition engine that decides which guests get recognized, when, and howbuild models that turn visit behavior, engagement, and referrals into personalized perks, content unlocks, and messagingown the data foundation: event schemas, feature pipelines, and the infrastructure that lets us learn from every guest interactionpartner with the founding team to define what signals matter — visits are just the startprototype and evaluate generative AI for guest messaging, with operator authenticity as the constraint, not an afterthoughtbuild evaluation frameworks that keep quality high as we scale from pilot venues to hundreds of operatorswork across POS data (square, toast), app behavior, and messaging engagement to build a unified guest graphship quickly, measure honestly, and retire models that don't earn their complexitypreferred qualifications4–8+ years of ML engineering experience; at least one tour at an early-stage startup or small teamyou've shipped production ML systems end-to-end: data, training, serving, monitoringstrong software engineering foundation — you can hold your own in any part of the stacktaste for knowing when ML actually helps vs. when heuristics or product design solve the problem bettercomfort with ambiguity — you can scope a problem, pick the simplest approach that works, and shipexperience working directly with founders or technical leadership in fast-moving environmentsbonus points for experience inrecommender systems, personalization, or ranking at consumer scalebuilding with LLMs in production: prompt engineering, evals, and guardrails for voice-sensitive use casesconsumer product companies where ML is a core part of the user experiencehospitality, retail, payments, or loyalty databuilding data and ML infrastructure from zeroevent-driven systems and real-time inference