Projects

AI Stock Market Predictor📈

  •  Built an interactive web app with Dash (Plotly) and Python using an LSTM deep learning model in TensorFlow and Keras to predict stock movements, achieving 90%+ accuracy on real-world financial data.
  • Developed a correlation matrix with Pandas and Matplotlib to analyze dependencies across 5+ companies, improving prediction reliability by 40% and identifying high-risk assets.

Tools: Python, TensorFlow, Keras, Pandas, Matplotlib. Dash (Plotly)

Link to project
GitHub Repository
Google Colab

Car Dealership Chatbot - Sponsored Winner🥳

  • Designed and deployed a live AI-powered chatbot using Natural Language Processing (NLP) with a React/Next.js frontend and Flask backend, reducing search time for users by 90% through personalized vehicle recommendations.
  • Built custom Named Entity Recognition (NER) and Sentiment Analysis (SA) models in Python, trained on 600+ automotive-specific phrases, enhancing intent recognition accuracy by 95%.

Tools: Python, SpaCy, Pandas, MongoDB, Flask, Matplotlib, React, Next.js, CSS

Link to project
Backend Repository
Frontend Repository

Robo Rover 🤖

  • Programmed autonomous navigation and targeting algorithms in Python on a Raspberry Pi, integrating wall-following, gyroscopic feedback, and color detection for accurate orientation and launching.
  • Designed and implemented motor control and sensor systems (ultrasonic, gyro, color) to enable obstacle avoidance, tunnel traversal, and precise alignment for ball launching.

  • Built a fully autonomous robotic launching system with custom motor control, reducing targeting error and enabling reliable completion of competition tasks under strict sensor/motor constraints.

RoboHacks 2024 🚗

  • Programmed ESP32 firmware in C++ with Wi-Fi connectivity for remote robot control, integrating Python scripts for wireless communication with a laptop.
  • Designed and implemented autonomous navigation across multi-terrain courses using servo and power motor control, enabling precise waste collection and deposition into categorized bins.
  • Built and tested a waste collection system with servo-controlled mechanisms for accurate sorting of recycling, compost, and garbage, improving scoring reliability under competition constraints.

Back to the top