Zen Iar

I build high-performance backend architectures, applied machine learning models, and generative AI systems. My work spans asynchronous REST APIs, private RAG pipelines, and production-ready cloud deployments.

I architected an enterprise e-commerce backend with FastAPI, PostgreSQL, SQLAlchemy, Alembic, Pydantic, JWT authentication, RBAC, Docker, and PyTest. I delivered modular asynchronous services, zero-downtime database migrations, and automated test coverage exceeding 90% of endpoints.

I trained a CNN retinal disease detection system that achieved 96.2% test accuracy across 4,000+ clinical images and exposed diagnostic inference through a FastAPI and OpenCV API. I also built a vehicle valuation and resale marketplace using regression and XGBoost approaches, reaching an R² score of 0.87.

I'm continuing my BSc. CSIT at Tribhuvan University while building reliable software and machine learning products that solve practical problems.