All Projects
Rapid Rescue AI
An intelligent emergency-response platform that uses multi-agent AI, machine learning, geospatial analysis, and real-time location intelligence to coordinate incident detection, emergency triage, ambulance dispatch, route optimization, and hospital recommendation. It supports dynamic routing, ETA prediction, emergency severity analysis, and offline-resilient communication.
Role: AI/ML, Deep Learning & Full Stack Development
Python, LangGraph, Multi-Agent Systems, FastAPI, Gemini, PyTorch, YOLO, OpenCV, XGBoost, React Native, TypeScript, Firebase, GPS/GNSS, A*, Dijkstra
TripMate AI
An MCP-based multi-agent travel planning system that coordinates Flight, Hotel, Weather, Budget, and Itinerary agents through real-time external tools. It dynamically routes requests using a Supervisor Agent and includes guardrails, human-in-the-loop approval, itinerary revision, checkpointing, and execution observability.
Role: Agentic AI Engineer
LangGraph, LangChain, LLaMA 3.3-70B, Groq, MCP, FastAPI, Python, PostgreSQL, Tavily, AviationStack, OpenWeatherMap, Docker, Render, LangSmith, Jinja2
QUBES
A quantum-secure real-time chat platform combining BB84 Quantum Key Distribution simulation with AES-256 encryption. The application supports real-time messaging, quantum key exchange, eavesdropper detection, encrypted image sharing, ephemeral session keys, on-demand rekeying, and secure ciphertext storage.
Role: Full Stack & Mobile Developer
React Native, React, TypeScript, Python, FastAPI, Firebase, Expo, BB84 QKD, AES-256, CryptoJS, AsyncStorage, Vercel, EAS Build
TRENDY
An all-in-one technology and creative ecosystem combining SaaS products, AI tools, digital services, events, and ticketing. The cross-platform application supports authentication, user profiles, event discovery, QR-enabled tickets, personalized dashboards, and reusable UI components across 20+ screens.
Role: Full Stack & Mobile Development
React Native, Expo, TypeScript, Expo Router, Firebase, Zustand, Reanimated, EAS Build, React, Node.js, Express, MongoDB, Tailwind CSS
Healthcare AI
An end-to-end healthcare platform integrating doctor discovery, appointments, online consultations, prescriptions, health records, medicine ordering, and Rapid Rescue AI emergency-response workflows. It includes role-based patient and doctor workflows, real-time communication, incident tracking, analytics, and medical document management.
Role: Full Stack Development
React, TypeScript, Vite, Tailwind CSS, shadcn/ui, Redux Toolkit, TanStack React Query, Firebase, Axios, Recharts, Vercel, WebRTC, Real-Time Communication (RTC),Peer-to-Peer Video Calling
Item-Based Book Recommendation System
A personalized book recommendation system built using item-based collaborative filtering and the Book-Crossing dataset. It preprocesses user-rating data, constructs a Book × User interaction matrix, calculates book-to-book similarity using Cosine Similarity, and recommends the Top 5 similar books.
Role: Machine Learning Engineer
Python, Pandas, NumPy, Scikit-learn, Collaborative Filtering, Cosine Similarity, Recommendation Systems, Data Preprocessing
Content-Based Movie Recommendation System
A personalized movie recommendation system that analyzes genres, keywords, plot overviews, cast, and directors. It uses feature engineering, text preprocessing, stemming, Bag-of-Words vectorization, and Cosine Similarity to generate the Top 5 most similar movies.
Role: Machine Learning Engineer
Python, NLP, CountVectorizer, Bag-of-Words, Stemming, Cosine Similarity, Content-Based Filtering, Feature Engineering
Popularity-Based Book Recommendation System
A book recommendation system that identifies highly rated and widely reviewed books using collective user behavior. It combines Books, Users, and Ratings datasets and uses rating frequency, average ratings, aggregation, and a 250+ rating threshold to generate the Top 50 popular books.
Role: Data Scientist & Machine Learning Engineer
Python, Pandas, Recommendation Systems, Data Analysis, Data Cleaning, Data Preprocessing, GroupBy, Aggregation, Statistical Analysis
Image Classification using CNN
An end-to-end Convolutional Neural Network for binary image classification using the Kaggle Dogs vs Cats dataset. The project includes image preprocessing, custom CNN architecture, model training, evaluation, and inference, achieving 93%+ training accuracy and 79.64% validation accuracy.
Role: Deep Learning Engineer
Python, TensorFlow, Keras, OpenCV, NumPy, Matplotlib, CNN, Computer Vision, Kagglehub, Google Colab
Hand Gesture Recognition
A real-time hand gesture recognition system using a CNN to classify six gesture classes: NONE, ONE, TWO, THREE, FOUR, and FIVE. It combines image augmentation and deep learning with OpenCV webcam processing, grayscale conversion, contour detection, convex hull, and convexity defects.
Role: Deep Learning Engineer
Python, TensorFlow, Keras, OpenCV, NumPy, Matplotlib, CNN, Computer Vision, Data Augmentation