Product Recommendation Data Scientist

MXT.co, a global remote-first Shopify Plus skincare ecommerce platform, is seeking a part-time Data Scientist to design and implement a sophisticated product recommendation engine. The ideal candidate will replace a legacy algorithm with a statistically rigorous, interpretable system that aligns quiz respondents to curated product selections, ensuring accuracy, scalability, and transparent explanations.

Job Overview

• Design and build a recommendation engine for product selection using historical quiz data.
• Develop explainable, deterministic models for five product categories.
• Integrate the recommender as a stateless API within Shopify Plus and custom quiz applications.
• Ensure recommendations comply with business logic, inventory, and category constraints.

Key Responsibilities

• Engineer weighted naive Bayes, regularized logistic regression, and kNN retrieval models.
• Optimize model ensembles via cross-validation to improve accuracy.
• Provide interpretable explanations for each recommendation.
• Integrate with external APIs (Shopify, Airtable) and support reliable production deployment.
• Monitor, log, and retrain the system for ongoing reliability and compliance.
• Prototype and optionally implement LLM-powered explanation agents.

Required Skills & Qualifications

• Proven experience in recommender systems using probabilistic models and classification methods.
• Expertise in Python, scikit-learn, pandas, numpy, and FastAPI or equivalent.
• Strong understanding of quiz-based personalization and multi-class classification.
• Ability to translate statistical results into clear, user-facing explanations.
• Experience working with low-sample datasets and advanced regularization techniques.
• Integration experience with production APIs and business logic enforcement.
• Knowledge of product taxonomies and category constraints.
• Bonus: Experience with LLMs, retrieval-augmented generation, and JSON schema enforcement.