Graph Architect

Role:
Knowledge Graph Architect

Location:

UK
Type:
Contract

Thebes Group is an optimisation company specialising in AI-enabled transformation. We help organisations improve workflow, reporting, information management and operational decision-making by combining process optimisation, knowledge architecture, semantic technologies, automation and artificial intelligence. We are currently delivering an AI transformation programme for a private equity group, focused on enhancing group-level operations through intelligent workflows, improved information accessibility, executive reporting and AI-driven operational intelligence.

A foundation ontology, taxonomy and initial knowledge graph exist. The data is mapped, manageable in scope and well understood within the team. An Ontology Engineer sits alongside this role to own the semantic foundation and an AI engineer handles the agent build. The Knowledge Graph Architect takes the semantic models produced by the Ontology Engineer and makes them operational: designing and running the graph platform, pipelines and integration layer that AI agents query and depend on.

The Role:

As Knowledge Graph Architect, you are responsible for the operational knowledge layer: the graph database, data pipelines, integration architecture and platform governance that sit between the semantic models and the AI agents consuming them. You will expand and maintain the existing knowledge graph as the programme evolves, ensure data flows correctly from source systems into the graph and work closely with the AI engineer to make the knowledge layer accessible, performant and reliable for agent use.

Where agent outputs are incorrect, you will work with the Ontology Engineer and AI engineer to identify whether the problem sits in graph structure, data integration or platform performance, and resolve it at source.

• Design and extend the enterprise knowledge graph architecture as operational requirements grow
• Implement and maintain graph database infrastructure using Neo4j or equivalent platforms
• Build and manage ETL/ELT pipelines that ingest group operational data into the knowledge graph accurately and consistently
• Design and optimise Cypher queries for agent consumption, analytics and operational reporting
• Integrate the knowledge graph with LLM and RAG architectures to support AI agent knowledge retrieval
• Implement Graph
  
  RAG patterns to enable agents to traverse and reason over graph-structured knowledge
• Ensure graph platform governance including security, access control, versioning and operational monitoring
• Work closely with the Ontology Engineer to ensure graph structures accurately reflect the semantic model
• Collaborate with the AI engineer to optimise how agents query and consume the knowledge graph
• Identify and resolve data integration issues that cause agent output failures or knowledge retrieval errors

• Neo4j (Enterprise)
• Amazon Neptune
• Stardog
• Graph
  
  DB (Ontotext)
• Tiger Graph
• Azure Cosmos DB (Gremlin API)
• Cypher
• SPARQL 1.1
• Gremlin
• GQL (ISO standard)
• open Cypher
• SPARQL Update
• Property graph modelling
• RDF graph modelling
• Labelled property graphs
• Hypergraph structures
• Entity resolution
• Graph schema design

• ETL/ELT pipeline design
• Apache Kafka
• Apache Spark
• AWS Glue
• dbt
• Apache Airflow
• Entity matching and resolution
• Data lineage tracking
• Schema mapping
• Semantic data integration
• Metadata management
• Data quality frameworks
• Python (networkx, rdflib, py2neo)
• SQL
• Bash/Shell Scripting
• REST API integration
• GraphQL
• JSON-LD processing

• RAG architecture design
• Graph
  
  RAG implementation
• Vector database integration
• LLM knowledge grounding
• Semantic retrieval design
• Agent knowledge API design
• AWS (Neptune, Glue, S3, Lambda)
• Azure (Cosmos DB, Synapse)
• GCP (Vertex AI, Big Query)
• Terraform/IaC
• Docker/Kubernetes
• CI/CD pipeline management
• Graph performance optimisation
• Access control and security
• Backup and recovery
• Monitoring and alerting
• Schema versioning
• Operational runbooks

• Proven experience designing and implementing knowledge graphs in a production or client-facing environment
• Hands‑on Neo4j or equivalent graph database capability including Cypher…