Final year student interested in machine learning, deep learning, and building things. Working on NLP, CV, and full-stack applications.
Coursework spanning statistics, machine learning, deep learning, algorithms, and application development. Tracks in NLP, computer vision, and generative AI.
End-to-end ML pipeline classifying 198,000 user comments into 4 categories (general, identity/hate, political, violent). Handled severe 20.9× class imbalance. Progressed from Logistic Regression → Random Forest → LightGBM with hyperparameter tuning. Built on Kaggle.
Classified audio into 10 genres from noisy mashups (ESC-50 noise overlay). Trained Simple CNN → ResNet-style CNN → AST (Audio Spectrogram Transformer). Final ensemble (AST × 0.75 + ResNet × 0.25) achieving ~0.85 val macro-F1. All runs tracked on W&B.
Web-based placement management system connecting students, companies, and admins. Role-based access, resume upload, drive approval workflow, application tracking. Built with Flask, SQLite, and Bootstrap. Course project for Modern Application Development.
Production-grade rebuild with Vue 3 frontend and Flask-RESTful backend. Added Redis caching, Celery async jobs, scheduled deadline reminders, CSV export with email alerts, token-based auth via Flask-Security, and a Pinia message store. 6 concurrent services.
Elected secretary of Namdapha House, the student community of IIT Madras BS Degree, representing 3900+ students across India. Organised academic support, events, hackathons, and cross-community initiatives.
↗ View docsManaged the digital presence and web infrastructure of Namdapha House. Maintained community platforms and supported online events for a house of thousands of students spread across India.
↗ View docsLooking for full-time ML / data science. Feel free to reach out.