I design and build end-to-end AI systems, with a focus on:
- Fine-tuning LLMs (Qwen, LLaMA) and vision models (DINOv2, Clay, Prithvi)
- Computer vision for satellite imagery — damage assessment & change detection
- Production-ready backends using FastAPI & PostgreSQL
- Model optimization for deployment, latency, and edge constraints
I prefer systems that ship, not just experiments that stay in notebooks.
AI / ML
PyTorch · TensorFlow · Hugging Face · scikit-learn · OpenCV
Generative AI
LLM Fine-tuning · RAG · LangChain · Embeddings . Chatbot
Backend & Data
FastAPI · Flask · PostgreSQL · Convex
Infra & Tooling
Git · Docker · Linux · Jupyter
Cloud AWS . GCP
- Satellite imagery analysis pipelines (pre/post-event comparison)
- Scalable AI APIs for real-world use cases
- Deepening expertise in transformers and practical MLOps


.png?raw=true)