On NeoCraft, I built a mobile-first storefront and engineered service logic in Go and Next.js over Supabase and Firebase. The platform supports catalogue and stock management, orders, reviews, and an authenticated admin dashboard with sales analytics.
I also designed and shipped NeoCraft’s sales assistant across Messenger and Instagram. It answers catalogue questions, captures orders from conversations, and uses database-grounded product information to keep prices and availability current.
For my undergraduate thesis, I developed a multimodal machine-learning framework for poultry health monitoring using images and vocalization signals. The image model reached 99.61% accuracy across four health conditions, while an RBF-SVM reached 98.33% mean accuracy on audio analysis.
My other projects include a Linux network namespace simulation, a Flask emotion-based music recommendation system, and an Android golf game. At HackTheAI, I built REST APIs, business logic, and database operations for an AI-powered tour guide platform.
