nDeliver the natural-language query capability over the Entity Graph. This role implements the vector indexing and retrieval layer that powers GraphRAG — enabling users to ask questions in plain language instead of writing graph queries — and is accountable for the accuracy, relevance and evaluation of those responses.
nKey Responsibilities
n
n
- Vector index design & build — design and implement the vector indexing strategy over graph projections and entity attributes, including chunking, embedding selection and index configuration.
n
- Embedding pipeline — build pipelines to generate, store and refresh embeddings as entity and graph data changes.
n
- GraphRAG implementation — combine vector similarity search with graph structure and multi-hop traversal to produce grounded, context-rich retrieval.
n
- Natural-language query enablement — implement and tune NLQ scenarios agreed with WK; support natural-language-to-graph-query translation approaches.
n
- Retrieval evaluation & tuning — define and run evaluation harnesses measuring retrieval relevance and answer quality; tune retrieval parameters against agreed scenarios.
n
- Grounding & traceability — ensure retrieved answers are attributable to source entities and edges, preserving provenance.
n
- Performance & cost management — optimise index size,
query latency and compute/token cost of retrieval operations.
n
- Documentation — document retrieval architecture, evaluation results, known limitations and supported query patterns.
n
n
nMust-Have Qualifications
n
n
- 5+ years engineering experience with 2+ years hands-on vector search / RAG implementation
n
- Demonstrable production experience building a retrieval pipeline (not prototype-only)
n
- Robust Python skills and familiarity with embedding models
n
- Experience evaluating and tuning retrieval quality with defined metrics
n
- Understanding of how to ground responses and preserve source traceability
n
nNice-to-Have
n
n
- Direct GraphRAG experience (graph + vector combined retrieval)
n
- Familiarity with Microsoft Fabric NL2GQL / Data Agent capabilities
n
- Exposure to graph databases and traversal concepts
n
- Experience managing LLM inference cost and latency at scale
n
nKey Deliverables Owned
n
n
- Vector index over graph projections
n
- Embedding generation and refresh pipeline
n
- GraphRAG retrieval capability supporting agreed NLQ scenarios
n
- Retrieval evaluation results and tuning documentation
n
- Documented supported query patterns and known limitations
n