Architect - Search Relevance

We're looking for an Architect (search relevance) to own the end-to-end search relevance stack for a high-scale fashion and lifestyle e-commerce platform serving millions of customers and a catalogue of 20M+ styles. This is a high-impact IC architecture role at the intersection of machine learning, NLP, search, GenAI, and large-scale distributed systems.

The candidate will have responsibilities across the following functions:

Search Intelligence and Relevance:

• Design and own query understanding: intent classification, category prediction, attribute extraction, query rewriting, and spell correction.
• Build and fine-tune LLM and NLP models for domain-specific query comprehension.
• Own autocomplete, personalized typeahead, trending, and session-aware suggestions.
• Drive hybrid search combining lexical and dense/semantic retrieval.
• Solve low-recall and zero-result queries through query expansion, taxonomy mapping, and semantic fallback.
• Establish rigorous evaluation using NDCG, MRR, offline judgment sets, and A/B experimentation.
• Rapidly evaluate emerging LLMs, RAG, and generative search techniques and take successful ideas to production.

Systems and Architecture:

• Own the technical architecture across retrieval, indexing, model serving, and online inference.
• Design model-serving infrastructure, including feature computation, model versioning, A/B routing, shadow deployments, and latency optimization.
• Build catalogue enrichment and signal-generation pipelines for retrieval and NLP models.
• Drive build-vs-buy decisions and establish engineering standards for production ML systems.

Requirements:

• 10-15 years of industry experience.
• 4+ years building production search relevance or recommendation systems at scale.
• Strong ML fundamentals across ranking, retrieval, classification, embeddings, and neural networks.
• Strong hands-on NLP/Transformer experience with BERT, T5 or similar.
• Experience with LLM-based query rewriting, RAG, or generative search.
• Conceptual understanding of Elasticsearch, Solr, or OpenSearch.
• Strong Python skills with the ability to review production code in Java/Scala or Go.
• Experience with distributed systems, data pipelines, and low-latency model serving.
• Hands-on exposure to model-serving infrastructure, online feature stores, and deployment pipelines.
• Strong experimentation and statistical evaluation mindset.

Good to Have:

• Hands-on experience with Elasticsearch/Solr/OpenSearch relevance tuning.
• Experience with Qdrant, Milvus, Vespa, FAISS, or other ANN systems.
• Knowledge of e-commerce taxonomy or knowledge graphs.
• Publications or applied research contributions in IR, NLP, or recommendation systems.