We are building an enterprise-grade AI platform to enable secure, scalable and production-ready use of Generative AI and ML across the Customs Declaration Service.
The Knowledge Engineer will build and operate the data pipelines that populate and maintain the CDS knowledge graph at scale. You will transform structured and unstructured source data into a governed, high-quality knowledge asset, working from the Knowledge Modeller’s schema and platform retrieval design to support RAG and agent-based AI capabilities.
Key Responsibilities
- Build and maintain ingestion pipelines that programmatically populate knowledge graphs from structured and unstructured data sources.
- Develop graph population workflows covering entity and relationship extraction, loading and validation against defined schemas.
- Work with graph databases and query technologies such as Neo4j/Cypher or RDF/SPARQL.
- Develop Python-based data engineering and ETL processes for high-volume ingestion.
- Define and maintain effective chunking and vector embedding strategies for content used in knowledge retrieval.
- Monitor and maintain knowledge graph data quality, freshness and lineage.
- Implement reconciliation processes to identify and address ingestion gaps, including missed webhook events, without placing additional load on source systems.
- Collaborate with AI Engineers to ensure the knowledge graph and retrieval layer effectively support RAG and agent-based use cases.
- Contribute to the evaluation of graph and retrieval quality, with a focus on data completeness and coverage.
Essential Skills & Experience
- Hands-on experience building and populating knowledge graphs programmatically, rather than through UI-based curation.
- Strong experience with Neo4j/Cypher, RDF/SPARQL or equivalent graph technologies.
- Strong Python data engineering experience, including ingestion and ETL at scale.
- Practical experience with vector embeddings and chunking strategies for RAG or knowledge retrieval.
- Experience working with structured and unstructured data sources.
Desirable
- Experience with LangGraph or equivalent AI orchestration frameworks.
- Exposure to AWS-native AI services such as Amazon Bedrock and OpenSearch.
- Experience working within regulated or government environments.