Designed and implemented a RAG-based chatbot in my website, utilizing GROQ as the LLM, PineCone DB for vector storage, and ’multilingual-e5-large’ as the embedding model, orchestrated via LangChain for efficiency.
StreetSpecter is an innovative solution that leverages advanced computer vision and deep Learning to automatically detect, classify, and report road damage, primarily focusing on potholes.
Developed a model to predict stock prices using advanced time series analysis techniques.
A real-time weapon detection web application using Django, with computer vision and deep learning (YOLOv8) to identify weapons like knives, pistols, and rifles. Integrated email notifications for client alerts.