Staff Applied ML/AI Scientist - Search

Faire•San Francisco, CA

1d•$231,000 - $318,000•Hybrid

About The Position

As a Staff Applied AI/ML Scientist on the Search Group, you’ll drive the technical vision, ML algorithm strategy, and system design powering one of the most critical levers for customer value and company growth—Search (think about what you do when you land on any e-commerce site). You’ll lead the advancement of real-time Search and Recommendation systems behind our next-generation shopping experiences. You’ll operate at the forefront of algorithms—combining large language models, natural language processing, query understanding, deep learning, transformer-based sequential modeling, graph neural networks, and structured behavioral data to return hyper-relevant, personalized products/brands for any given query from the users. This is a rare opportunity to own end-to-end personalization in a high-scale, deeply multi-modal environment—while mentoring a team of talented scientists and engineers.

Requirements

7+ years of experience building large-scale ML systems, including 3+ years in search, recommendation, or ads ranking.
Hands-on experience with deep learning libraries (e.g. PyTorch) and vector search infrastructure (e.g. Faiss, ScaNN, Pinecone).
A strong track record of productionizing models that blend LLMs (e.g. BERT, GPT-class) with structured features to drive personalization.
A product-focused mindset and a bias toward execution—you move quickly from paper to prototype to production.
Strong Python skills, deep respect for system reliability and ownership, and experience operating in high-stakes environments.
Excellent communication and cross-functional influence—you raise the technical bar beyond your immediate team.

Nice To Haves

Contributions to open-source ML libraries or peer-reviewed publications in ML/AI.
MS or PhD in Computer Science, Statistics, or a related STEM field.
Strong practices around model development, agent workflow evaluation, and MLOps.

Responsibilities

Own the next-generation Search engine, integrating LLMs, query understanding, dense vector retrieval, deep personalization embeddings, multi-stage ranking, and reinforcement learning to serve personalized product feeds with <100ms latency.
Design and productionize natural language search and discovery systems, enabling intelligent agents to generate relevant and personalized collections, explain search results, and assist retailers in browsing, filtering, and evaluation.
Lead model development and GPU-based deployment efforts, leveraging frameworks like Triton to scale inference reliably and efficiently.
Mentor and grow senior Applied Scientists and MLEs, and establish best practices around model development, agent workflow evaluation, and MLOps.