Data Science Intern

Faire•San Francisco, CA

7d•Hybrid

About The Position

Faire leverages machine learning and data insights to transform the wholesale industry, giving independent retailers the tools to compete with large-scale e-commerce platforms and big-box stores. Our Data Science team builds and maintains the algorithmic systems — spanning search, personalization, recommendation, and ranking — that power our marketplace and help our customers thrive. We are hiring Data Science interns across several teams and are looking for intellectually curious, self-directed problem solvers eager to work end-to-end on high-impact challenges, from data exploration to production-ready solutions. Our internships are paid, 12–14 weeks in duration, with flexible start dates. Extensions are considered based on project scope and mutual interest.

Requirements

Currently enrolled or recently graduated Master's or PhD students in Computer Science, Operations Research, Statistics, Econometrics, or a related technical discipline
Publications or submissions to top-tier venues such as KDD, RecSys, ICML, NeurIPS, WWW, or SIGIR
Experience with recommender systems (collaborative filtering, deep recommenders, ranking), representation learning and embeddings, sequential models (RNNs, Transformers for user behavior modeling), bandit and reinforcement learning methods, and large-scale retrieval and ranking systems
Familiarity with offline evaluation metrics (NDCG, MAP, recall) and online experimentation
Experience working with large-scale or production datasets

Responsibilities

Design, develop, and A/B test cutting-edge machine learning algorithms and analytical solutions, with guidance from senior technical leads
Communicate project objectives, methodologies, and results clearly to both immediate teammates and broader cross-functional stakeholders
Navigate the complexity of a two-sided marketplace, identifying and addressing the unique challenges that arise at the intersection of retailer and brand needs
Design and deploy state-of-the-art recommender systems that power ranking and discovery across the marketplace
Develop rich user and item representations through embeddings, sequence models, and graph-based methods
Build real-time and streaming data pipelines that enable dynamic, context-aware personalization at scale
Apply exploration–exploitation strategies — including contextual bandits and reinforcement learning — to optimize recommendations under uncertainty
Advance recommendation quality through improvements to diversification, novelty, and long-term user engagement
Own the full ML lifecycle: from problem formulation and modeling through offline evaluation and online experimentation