RAG & Data Infrastructure Engineer

Role overview Build the data backbone that lets an organization put its private knowledge to work. The role centers on designing and operating ingestion and retrieval pipelines that connect unstructured company information to the people and language models that need it, with security and reliability treated as first-class concerns rather than afterthoughts.

Responsibilities - Design and maintain ingestion pipelines that pull documents, conversations, and other internal artifacts into a clean, queryable form. - Build retrieval systems (vector indexes, hybrid search, metadata filters) tuned for both human queries and model prompts. - Implement access controls, audit logging, and tenant isolation so sensitive material stays private. - Monitor retrieval quality, latency, and freshness, then iterate on embeddings, chunking, and re-ranking based on real usage. - Partner with applied engineers to wire retrieval into downstream products and agents.

Requirements - Hands-on experience building production RAG or semantic search systems. - Comfort with vector databases (such as pgvector, Pinecone, Weaviate, or Qdrant) and traditional search stacks like Elasticsearch or OpenSearch. - Strong Python skills and familiarity with embedding models, chunking strategies, and evaluation harnesses. - Solid grasp of security patterns for internal data: identity-aware access, encryption, secrets management, and least-privilege design. - Ability to debug end-to-end, from raw source documents to a retrieved chunk landing in a model context window.

Nice to have - Experience with knowledge-graph or structured-retrieval approaches alongside vector search. - Familiarity with orchestration frameworks such as LangChain, LlamaIndex, or custom pipelines.