Principal Machine Learning Engineer- Search Quality

Snowflake•Menlo Park, CA

About The Position

Snowflake is about empowering enterprises to achieve their full potential — and people too. With a culture that’s all in on impact, innovation, and collaboration, Snowflake is the sweet spot for building big, moving fast, and taking technology — and careers — to the next level. Snowflake is about empowering enterprises to achieve their full potential — and people too. With a culture that’s all in on impact, innovation, and collaboration, Snowflake is the sweet spot for building big, moving fast, and taking technology — and careers — to the next level. As Snowflake expands its product ecosystem, the ability for users to find relevant information across a fragmented landscape of data and metadata is paramount. The Snowscope team is at the heart of this mission, building and maintaining the internal search system that powers discovery across diverse corpuses, including the Catalog, Marketplace, Documentation, Workspaces, Notebooks, and more. We also maintain Universal Search, providing a seamless, single-entry search experience across all categories We are looking for a Principal Software Engineer to serve as the technical leader for Search Quality. This individual will be responsible for transforming how we measure and improve search relevance, moving from heuristic-based approaches to a disciplined, data-driven framework. You will identify key areas of investment, bridge the gap between traditional search and modern AI, and ensure that our search technology is ready for the next generation of AI-driven agentic workflows. OUR IDEAL PRINCIPAL SOFTWARE ENGINEER WILL HAVE:

Requirements

15+ years of industry experience designing, building and supporting large scale distributed services.
Has built and optimized search systems at Snowflake-scale or equivalent high-growth environments.
Possesses a startup mindset, acting with urgency to deliver incremental improvements while building toward a long-term vision.
Is a subject matter expert in the latest developments in NLP, LLMs, and their application to Information Retrieval.
Search Domain Expertise: Deep, hands-on experience with search technologies (e.g., Lucene/Elasticsearch/OpenSearch, vector databases) and a proven track record of improving search relevance and ranking at scale.
Deep ML Expertise: Extensive experience in machine learning specifically applied to search quality, including Learning to Rank (LTR), query understanding, and personalized ranking.
Hybrid Search Techniques: Intimate familiarity with blending semantic (vector-based, embeddings) and syntactic search (keyword-based, BM25) to achieve state-of-the-art retrieval accuracy.
Data-Driven Leadership: Ability to build a disciplined approach to search quality, including the design of evaluation frameworks (e.g., NDCG, MRR), A/B testing methodologies, and human-in-the-loop evaluation pipelines.
Technical Visionary: Demonstrated ability to translate high-level product goals into technical roadmaps and influence engineering teams to execute on a unified vision for Universal Search.
AI Agentic Frameworks: A forward-looking understanding of how traditional search systems must evolve to support AI agents, specifically focusing on RAG (Retrieval-Augmented Generation) and tool-use retrieval.
Distributed Systems: Strong foundation in building and scaling high-performance distributed systems that serve low-latency search results across massive, heterogeneous datasets.
Cross-Functional Collaboration: Proven ability to partner with and influence, Product Management and Data Science and AI team to define quality metrics and align technical investments with business impact.

Nice To Haves

Multi-Modal Search: Experience with multi-modal search (text, image, code) and understanding of how different corpuses (like Notebooks vs. Documentation) require specialized retrieval strategies.
Open Source Contribution: Active contributions to the search or ML open-source community.
User Experience Empathy: A strong sense of how search quality directly impacts the end-user experience and the ability to advocate for the user in architectural decisions.