Job Description

Role Description

We are seeking a data scientist to design, build, and integrate a new product recommendation engine for MXT.co, a Shopify Plus-based skincare ecommerce platform. The ideal candidate will create a deterministic, interpretable recommender system that aligns new quiz respondents with curated product selections based on historical quiz data, while improving accuracy, scalability, and explainability.

Responsibilities

• Design and implement a recommendation engine that predicts five product categories (Primary 1–3, Upsell 1–2) for each user based on quiz answers.
• Replace the current “1-answer = 1-vote” algorithm with a statistically rigorous, explainable, and production-ready recommender.
• Engineer multiple components: weighted naive Bayes, regularized logistic regression models, and kNN retrieval.
• Ensemble and optimize these components using cross-validation to improve match accuracy against seed-labeled data.
• Package the recommender as a stateless API and integrate it into a Shopify Plus storefront and custom quiz app, syncing with Airtable.
• Provide interpretable “why we recommended this” explanations per product category.
• Build and monitor logging, retraining, and guardrail systems to support reliability, category constraints, and business logic (e.g., inventory, pricing).
• Optionally prototype and integrate an LLM-powered agent to augment or generate recommendation explanations.
  
  

Requirements

• Strong experience building recommender systems using probabilistic models, logistic regression, or nearest-neighbor methods.
• Proficiency in Python and relevant ML libraries (e.g., scikit-learn, pandas, numpy); experience with FastAPI or similar frameworks.
• Familiarity with quiz or survey-based systems, personalization, and constrained multi-class classification.
• Ability to translate statistical outputs into human-readable product explanations.
• Comfortable working with low-sample datasets and applying regularization, smoothing, and tie-breaking techniques.
• Experience integrating ML systems into production environments with business logic and external APIs (Shopify, Airtable).
• Understanding of product taxonomies, category constraints, and optional clinical/business rule enforcement.
• Bonus: Experience with LLM-based retrieval-augmented generation (RAG), JSON schema enforcement, and deterministic prompting.

• للتقديم الان