Zylo is an intelligent note-taking companion designed to understand, connect, and expand your ideas. It offers intuitive daily journaling with calendar navigation and auto-save, smart link expansion that extracts key information and provides rich inline previews, and intelligent search with natural language queries and semantic understanding. Powered by RAG (Retrieval-Augmented Generation), Zylo transforms your notes into a dynamic knowledge base, enabling cross-note connections, intelligent summarization, and knowledge synthesis. It provides secure Google Sign-in and is built for an AI-first experience, moving beyond traditional basic note apps to amplify your thinking.

Created an AI-powered chatbot that translates natural language queries into SQL statements based on database schema, retrieving and presenting information in natural language. Integrated defog-sqlcoder-7b-2 model for state-of-the-art natural language to SQL generation with high accuracy in complex query scenarios. Designed with separation of concerns between query generation, schema interpretation, and response rendering. Applied defensive programming techniques and async-safe WebSocket handlers for high availability. Configured structured logging and error handling for production-readiness.

Developed a PyTorch-based application that colorizes grayscale images using pre-trained models (ecvv16 and siggraph). The project features a Streamlit web interface allowing users to upload grayscale images and receive colorized versions in real-time. It integrates deep learning models for image colorization and provides an intuitive user experience for visualizing results.

Designed and developed StreetSpecter, an AI-driven system for real-time pothole detection, aiming to improve road safety and maintenance efficiency. Custom-trained a YOLOv8s model on 5,493 images over 25 epochs using a Colab T4 GPU, achieving 90% detection accuracy. Integrated OpenStreetMap API to map over 1,000 km of road data, enabling geolocation of detected potholes and contributing to a potential 30% reduction in maintenance response times. Adopted best practices like modular Flask blueprints, route-based caching, and async-compatible I/O for OpenStreetMap data. Ensured reproducibility using configuration-driven training and automatic data augmentation pipelines. Deployed using Docker for consistent environment control across systems.