Senior AI Engineer (Search/Retrieval)
Core
Build the retrieval layer that powers enterprise AI agents by stitching together data from heterogeneous sources (Salesforce, Zendesk, Jira, etc.) into ranked, permission-aware responses.
Role type
Senior IC machine-learning engineer (search/retrieval)
Builds
Unified retrieval layer, hybrid retrieval pipelines, ingestion/freshness pipelines, permission-aware retrieval engine, query understanding modules, chunking/embedding strategies, evaluation harnesses, production-grade observable systems.
Domain
Enterprise SaaS, AI Agents, Information Retrieval, Knowledge Graphs
Deliverable
production ML models | product features
Required skills
Production search/retrieval system design, hybrid retrieval (lexical/dense/structured), vector search, knowledge graphs, LLM-driven reasoning, permission-aware data access, incremental data ingestion, query understanding, intent parsing, entity linking, LLM-assisted query rewriting, chunking strategies, embedding strategies, evaluation metrics (NDCG, MRR, recall, faithfulness), system observability, Python/Go/Java/Backend development
Preferred skills
Experience with enterprise systems (Salesforce, Zendesk, Jira, SharePoint), experience with agentic workflows, experience with MCP (Model Context Protocol)
Technologies
Python, Go, Java, BM25, Vector Search, SQL, Knowledge Graphs, LLMs, NDCG, MRR, Recall@k
Responsibilities
Build a unified retrieval layer across enterprise systems; Design hybrid retrieval pipelines combining lexical, dense vector, and structured retrieval; Engineer ingestion and freshness pipelines for millions of records; Own permission-aware retrieval (ACL preservation); Build query understanding for agents (intent parsing, entity linking, query rewriting); Design chunking and embedding strategies for diverse content types; Build evaluation and experimentation harnesses; Ship production-grade, observable systems with strong SLOs; Mentor teammates on retrieval architecture and engineering craft.
Seniority
Senior, hands-on IC