Role overview
RAG / Vector Database Engineer
EngineeringOpen
Own retrieval infrastructure so documentation, subnet metadata, and protocol context surface accurately in product flows.
Remote · Full-time preferred · Contract for scoped work
Outcomes
- Design embedding pipelines, chunking strategies, and refresh jobs for evolving subnet and docs corpora.
- Operate vector stores with filtering, hybrid search, and cost-aware indexing at production scale.
- Measure retrieval quality (recall, latency, hallucination rate) and iterate with product and agent teams.
Baseline
- ·Hands-on experience with vector databases and RAG systems in production.
- ·Solid understanding of embeddings, re-ranking, and metadata-aware retrieval patterns.
- ·Comfort owning data hygiene: deduplication, versioning, and PII or secrets exclusion.
Plus:
Background with Pinecone, pgvector, Qdrant, Weaviate, or similar stacks.