Briju Demel

At Deloitte, I build GenAI RAG infrastructure and agentic AI systems for Dell Technologies, turning unstructured data into structured, queryable knowledge across vector and graph databases.

I've built FastAPI services for data ingestion, vector, graph, full-text, and hybrid search pipelines, and retrieval-quality evaluation. I also built RAGForge, a platform for ingesting files from any source and querying vector and graph stores for downstream agents.

I designed and deployed a LangGraph-based Dell Services Troubleshooting Agent to triage hardware-related internal queries, with validation checks using open-source frameworks. I've also added persistent agent memory with langmem and mem0 and integrated LLM tools that generate graph-search queries from uploaded documents.

Independently, I built Codebase Assistant, which indexes repositories through AST parsing with tree-sitter and Neo4j dependency graphs to answer natural-language questions about arbitrary codebases. Earlier, I developed Next.js applications for startups at Kylo Apps, including authentication, payment portals, and Google Analytics work.