Principal Engineer - RAG Database & Embeddings Architect
Core
Designing, building, and governing the vector database and retrieval architecture that powers enterprise Retrieval-Augmented Generation (RAG) systems.
Role type
Principal Engineer - RAG Database & Embeddings Architect
Builds
Enterprise RAG data stores, vector databases, document stores, metadata stores, and hybrid search layers.
Domain
Financial services / AI / Vector Search / RAG
Deliverable
production ML models
Required skills
Vector databases, embeddings, similarity search, approximate nearest neighbor algorithms, retrieval optimization, RAG architectures, distributed systems, database design, API design, performance tuning, metadata modeling, access control filtering, document provenance, auditability, engineering patterns, standards definition, cross-team architecture leadership, technical strategy, mentoring
Preferred skills
Hybrid retrieval, reranking models, knowledge graphs, entity-aware retrieval, LLM evaluation, retrieval evaluation, automated relevance testing, model drift management, embedding drift management, re-indexing strategies, high-quality retrieval architecture, scalable vector database strategy, enterprise architecture, cloud-native engineering, observability, technical roadmap development, cross-functional influence, vendor evaluation, production support
Technologies
Azure AI Search, Cosmos DB vector search, Pinecone, Weaviate, Milvus, OpenSearch, Elasticsearch, PostgreSQL/pgvector, OpenAI, Azure OpenAI, Cohere, Hugging Face
Responsibilities
Define engineering approach for program/portfolio solutions; lead planning and design of complex features spanning multiple teams; create ideas for complex technology and solution development approaches; lead technical oversight including design reviews and code within own domain; define technology tool stack; explore state-of-the-art technologies to improve development efficiencies; lead end-to-end test strategy and integration; improve developer experience; advance technology platforms through innovation; reduce risk and improve quality across technology portfolio; design and own architecture for enterprise RAG data stores; define embedding strategies across structured, semi-structured, and unstructured content; evaluate and select embedding models; design vectorization workflows including chunking, embedding generation, indexing, versioning, and re-embedding lifecycle management; implement semantic, keyword, metadata-filtered, and hybrid retrieval patterns; optimize retrieval quality using similarity metrics, reranking, query expansion, metadata boosting, and relevance feedback; establish standards for vector schema design, namespace strategy, document lineage, source attribution, and access-control-aware retrieval; partner with data pipeline engineers to ensure high-quality ingestion; partner with context engineers to tune retrieval outputs; define observability for retrieval quality; lead technical evaluation of vector database platforms and retrieval frameworks; provide engineering leadership, design reviews, mentoring, and architectural guidance; serve as senior technical authority for enterprise AI platform engineering; own architecture decisions impacting multiple teams; create reusable patterns, reference architectures, standards, and engineering guardrails; mentor senior engineers and influence technical direction; balance innovation with operational reliability, security, compliance, scalability, and cost management; communicate complex AI and data engineering concepts to stakeholders
Seniority
Principal, strategy & mentorship
