Computer Science Student · Deep Learning Enthusiast · Problem Solver
- 🎓 Pursuing B.Tech in Computer Science at Kalasalingam Academy of Research and Education (2022–2026)
- 🤖 Passionate about Deep Learning, AutoML, and AI-powered applications
- 🐦 Built CNN models for real-world problems — from bird sound classification to cancer detection
- 🏆 Hackathon competitor with wins at Track Tracer, WEBX, and experience at Intel AI Hackathon & G2-HackFest
- 📍 Based in Karaikkudi, Tamil Nadu, India
- 💬 Languages: English · Tamil · Hindi
Languages
Frameworks & Libraries
Tools & Databases
CNN model achieving 78.75% accuracy classifying 114 bird species from audio recordings using MFCC feature extraction — deployed via Streamlit for real-time predictions.
Python TensorFlow Keras OpenCV Streamlit
AutoML pipeline using Differentiable Architecture Search (DARTS) to design optimized CNN architectures that outperform ResNet baselines in accuracy and efficiency on the SIPaKMeD dataset.
Python PyTorch AutoML scikit-learn Matplotlib
Neural Architecture Search system classifying lung-colon cancer from histopathology images with 99.72% accuracy across 5 classes.
Python PyTorch NAS Deep Learning
Deep learning model that detects facial expressions in real time and curates a music playlist based on the user's detected emotion.
Python TensorFlow Keras OpenCV CNN
Conversational AI tool built with the Gemini API to assist parents of autistic children with empathetic, tailored guidance on therapy, communication, and sensory needs.
Python Gemini API Flask HTML/CSS/JavaScript
AI-powered failure prediction platform using BiLSTM neural networks with 97.4% accuracy and 6-hour failure forecasting via a live WebSocket-streaming dashboard.
Python BiLSTM WebSocket Deep Learning
"Passionate about building scalable, efficient applications and solving real-world problems through clean, effective code."
⭐ If you find my work interesting, feel free to star a repo or connect with me!