AI-focused CS student building full-stack products, scalable APIs, RAG pipelines, & intelligent AI-powered applications

github

Manash Khandelwal  • Peerlist

This project introduces a Grocery Store Object Detector, developed using YOLOv8s, designed to identify grocery store items in real-time. The model detects 24 predefined grocery products using a robust object detection framework. It is hosted locally via a Flask-based web interface, where users can upload images to detect and return identified objects. The project aims to simplify the checkout process in grocery stores, potentially allowing store employees to scan and bill entire carts efficiently using their smartphones. The dataset for this project was sourced from Roboflow, with extensive training and testing to ensure accuracy. Experimental results validate the model's capability to identify products effectively, laying the foundation for future deployment in real-world retail environments.

Grocery-Store-Object-Detector

Vidhi AI is an innovative legal tech platform combining AI-powered assistance with real lawyer consultations. Built on Mistral 7B with RAG technology, it provides instant legal information through an intelligent chatbot while seamlessly connecting users with verified lawyers when expert advice is needed. The system addresses India's legal accessibility challenges through a React/Node.js web application with secure authentication, payment integration, and real-time chat functionality. Over 12 weeks of agile development, the team implemented core features including AI legal query processing, lawyer booking system, wallet-based payments, and user feedback mechanisms. Vidhi AI demonstrates strong market potential with its unique hybrid model and pay-per-consultation monetization strategy.