Staff Software Engineer – Search Quality

About the role

As a Search Quality Engineer at Databricks you will own the relevance of results for both AI agents and human users. You will shape the retrieval backbone that powers large language models to access up‑to‑date enterprise data and ensure employees can find assets through high‑recall search interfaces.

Key responsibilities

• Design and maintain hybrid retrieval systems that combine keyword search with semantic vector search.
• Develop and tune ranking models to satisfy both human readability and LLM‑ready context.
• Build evaluation frameworks and relevance metrics (nDCG, MRR, Precision@K) to continuously monitor search quality.
• Integrate heterogeneous data sources, including SQL tables, unstructured documents, and real‑time business metrics.

Required profile

• Deep experience with information retrieval technologies such as Lucene or Elasticsearch.
• Strong focus on relevance measurement and metrics, building human‑in‑the‑loop and LLM‑based evaluation pipelines.
• Ability to bridge classic IR techniques with modern Retrieval‑Augmented Generation (RAG) approaches.

Required skills

• Lucene
• Elasticsearch
• Embeddings
• Ranking algorithms