About this role
WSP seeks a Knowledge Engineer for its Engineer - AI position to support MAV’s structured ingestion, reasoning and retrieval of complex mining and asset information. The role turns unstructured and semi-structured technical documents into RDF knowledge graphs and develops graph-native retrieval workflows, including GraphRAG. The position is listed in Noida and Mumbai, India, and WSP describes a hybrid work model.
Responsibilities
- Design automated extraction pipelines for tailings storage facility documents, including PDFs, scanned reports, tables and figures. Use LLM- and vision-based methods such as OCR, layout understanding and multimodal extraction to identify entities, attributes, relationships and supporting evidence; validate and normalize the results.
- Design and evolve ontologies and schemas for tailings storage facility, risk, asset and operational data. Define RDF/OWL classes, properties, constraints and semantic relationships, aligning them with industry standards and MAV data models.
- Build RDF knowledge graphs in graph repositories and map extracted information to graph structures with source-document traceability, versioning, provenance and evidence links.
- Develop retrieval pipelines using SPARQL, reasoning and, where appropriate, embeddings. Build GraphRAG workflows based on structured graph context and enable natural-language queries for AI assistants and analytics tools.
- Collaborate with developers, domain experts and AI engineers on extraction logic and schemas, and support integration with cloud AI and document-processing services.
Qualifications
- Proven experience in knowledge, ontology or knowledge graph engineering; strong knowledge of RDF, OWL, SPARQL and semantic data modeling.
- Hands-on experience with RDF graph repositories such as GraphDB, RDFox, Apache Jena or Neptune; experience building LLM-based extraction and structured-graph retrieval pipelines.
- Familiarity with OCR, layout analysis, multimodal document extraction, Azure-based AI and data platforms, and GraphRAG or hybrid graph-and-LLM architectures.
Employment with WSP India is subject to successful third-party background verification.
Skills for this role
Knowledge engineeringKnowledge extractionOntology designKnowledge graphsRDFOWLSPARQLSemantic data modelingGraphDBRDFoxApache JenaAmazon NeptuneGraphRAGLarge language modelsOCRLayout analysisMultimodal extractionEmbeddingsGraph reasoningAzure AIDocument processingData provenance