I build agents that do the work, APIs that hold up in production, and ML that makes it out of the notebook.
- 🤖 Agentic AI systems and RAG pipelines, built on Python backends
- ⚙️ Async FastAPI, JWT and RBAC, Redis, Celery, Alembic, Docker
- 🚀 Models shipped as real services, not notebooks: Streamlit and FastAPI, containerised
- 💬 Ask me about agents, API design, or getting a model from notebook to deployed
computer-agent An agent that sees your screen and controls your computer autonomously, using Claude vision.
astats Agentic AI system for applied statistical workflows, from data intake through analysis.
advanced-backend Task management REST API covering the hard parts: async FastAPI, JWT and RBAC, Redis caching, Celery, WebSockets, soft deletes, Alembic migrations, Docker.
ai-rag-api Production-grade document intelligence API. FastAPI, ChromaDB, Claude, async SQLAlchemy 2.0, pytest, Docker.
Expense_tracking_system Full-stack expense tracker. FastAPI, MySQL, Streamlit, Pandas, pytest, structured logging.
Car_Damage_Detection Vehicle damage classifier across 6 classes. ResNet50 transfer learning tuned with Optuna, served through both Streamlit and FastAPI.
credit-risk-modelling End-to-end credit risk scoring. SMOTETomek + Optuna, AUC 0.98, deployed as a scoring app.