Knowledge Graph & GraphRAG Lead | GenAI
Location: London | UK (hybrid) |
Type: Permanent or contract-to-perm
Own the enterprise knowledge graph and GraphRAG platform powering search, Q&A, and agentic apps.
ROLE PROFILE:
* Model the ontology/schema and build governed ingestion (NER/NEL, dedupe/merge, provenance).
* Deliver hybrid retrieval (graph traversal + embeddings [+ rerank]) with measured uplift.
* Expose secure APIs/tools for LLMs/agents; enforce access controls, rate-limits, auditability.
* Embed evals & observability; ship via CI/CD.
Must-have
* Shipped knowledge graphs in production (Neo4j/TigerGraph/Neptune or RDF/OWL; Cypher/SPARQL/GSQL).
* Delivered GraphRAG with evidence it beats vector-only.
* Python + LangChain/LlamaIndex; vector stores (Pinecone/Chroma/LanceDB).
* Cloud delivery (AWS/Azure/GCP) with basic IaC/CI/CD and governance.
Nice-to-have
* Text-to-Cypher/SPARQL with safety filters and small eval sets.
* MCP-style tool contracts for safe agent access.
* Streaming/ELT at scale (Kafka/Databricks/PySpark).