Agentic RAG Advisor for Health & Culture Tourism
RAG assistant with intent classification, hybrid retrieval, retrieval grading and re-query.
- Python
- LangGraph
- vLLM
- Qdrant
- 5 more
Role
AI Engineer
I take intelligent systems from research to a running product — and I’m open about the AI tools that help me build them.
Explore ↓I’m an AI engineer at Panates and a final-year M.Sc. student in Computer Engineering at Fırat University. I build end-to-end LLM systems — agentic RAG assistants, natural-language-to-SQL agents, and fine-tuned models served on self-hosted infrastructure — and I’ve written and defended national and EU R&D proposals along the way. My thesis looks at small language models collaborating on edge devices for privacy-preserving medical diagnosis.
Case studies from research question to running system. Each one reads complete on its own.
RAG assistant with intent classification, hybrid retrieval, retrieval grading and re-query.
Legislation-grounded Q&A with query expansion, multi-stage retrieval and reranking.
Metadata-first NL→SQL agent: plan, synthesise, validate and re-plan across SQL dialects.
Small demos of my work will run here, on my own server. Until then, the case studies explain each system.
Coming soon
Read the case studies →Notes on research, evaluation and building with language models.
First posts are on the way.
Questions about a project, a paper or a role — write to me.