Software Engineer
Capgemini
Building an end-to-end knowledge-graph pipeline that transforms PDF documentation into queryable semantic data.
- Developed a Python/FastAPI backend generating ontology and knowledge-graph Turtle (.ttl) files from PDF documents across two distinct processing groups
- Industrialized the full pipeline with structured error handling, centralized logging, and environment-variable-driven configuration
- Integrated GraphDB import workflows, vector store initialization, and semantic QA queries with persistent result storage
- Built a Vite + React dashboard to orchestrate uploads, trigger pipeline runs, monitor logs, and download artifacts
- Implemented interactive graph visualization using ReactFlow for intuitive ontology exploration
- Continuously improved cross-component compatibility for seamless dashboard integration
