AI/ML enthusiast with expertise in building end-to-end solutions, passionate about innovation, problem-solving, and creating impactful, user-centric applications.
Focused on developing intelligent systems that combine deep learning with retrieval augmentation. Currently building production-ready RAG pipelines with LangChain, ChromaDB, and embedding models to help engineering teams reduce development time.
- Loopot - Multi-agent property assistant combining vision-based issue detection and tenancy support
- Fracture Detection System - CNN-based detection of bone fractures in X-ray images with 94.56% accuracy
- Article Sentiment Analysis - NLP system for extracting emotional tones from articles with 86% accuracy
- Motorcycle Helmet Detection - Real-time helmet rule violation detection using YOLOv8 and Jetson Nano
