I build AI products end to end, from agent design and retrieval to the APIs and interfaces people use. My work focuses on agentic systems, MCP integrations and LLM tracing.
Production-RAG-MCP
RAG service with document ingestion, semantic search, a web chat widget and an MCP endpoint.
FastAPI · ChromaDB · Sentence Transformers · Docker
Medical-Bot
RAG chatbot for medical documents, working with Gemini and OpenAI models.
Python · Docker · AWS
gemini-rag-ragas-eval
RAG pipeline measured with Ragas evaluation.
Gemini · LangChain · FAISS · Ragas
YouTube Comments Sentiment Analysis
Sentiment pipeline from data collection to tracked experiments.
Python · MLflow · Docker
microgpt
Small GPT model, following Andrej Karpathy's microgpt.
Python
| Area | Tools |
|---|---|
| Languages | Python, TypeScript, JavaScript, SQL, Bash |
| AI and agents | LangChain, LangSmith, MCP, FastMCP, RAG, Hugging Face |
| Machine learning | PyTorch, scikit-learn, pandas, NumPy |
| MLOps | Grafana, MLflow, DVC, ZenML |
| Back end | FastAPI, Flask, Node.js, npm, Pydantic, asyncio, PostgreSQL, MongoDB |
| Front end | React, HTML, CSS, Figma |
| Infrastructure | Docker, Kubernetes, Azure, Keycloak, Git, Bitbucket, Linux |
| 3D and media | Blender, ComfyUI, Cinema 4D, Maya, ZBrush, After Effects |
