I'm a |
I am a MSc Computational Data Science student at Khalifa University, with a passion for machine learning and data analysis. I love turning complex data into clear insights and building user-friendly web applications.
I am pursuing my MSc in Computational Data Science at Khalifa University. My hands-on experience spans data analysis, machine learning model building, and front-end development. I enjoy turning raw data into clear insights and building user-friendly tools that solve real problems.
Based in Abu Dhabi, UAE • 056 697 2151 • English, Malayalam, Hindi, Arabic, Tamil
I keep my stack simple and reliable: strong web fundamentals and practical data tooling. Below are my core tools and how confident I feel using them.
Trained a BERT-based sentiment model for transliterated (Romanized) Malayalam text and deployed it as a production-style inference API. Built a FastAPI service with health checks, containerized with Docker, added automated tests + GitHub Actions CI, and deployed on Kubernetes (Minikube) with Service + Ingress.
Designed and trained a convolutional Variational Autoencoder (VAE) on 1024×1024 chest X-ray images. Extracted 64-dimensional latent embeddings and used them for multi-label disease classification. Evaluated performance using macro F1-score across pathologies.
Built an unsupervised anomaly detection system for MRI scans using a convolutional autoencoder. Trained the model to reconstruct healthy brain images and identified tumors via reconstruction error. Designed the system for anomaly localization without labeled tumor masks.
Built an NLP pipeline for Romanized Malayalam reviews: cleaned text, normalized transliteration, extracted features (TF-IDF, n-grams), and compared ML models (LogReg, SVM, NB, RF). Tuned with cross-validation; evaluated on precision/recall/F1.
Responsive portfolio with light/dark mode, ripple transition, language-aware logo typewriter, AOS animations, Netlify-ready deploy, and contact mailto/Netlify Forms fallback.
AI system to generate custom music using GANs. Role: model experimentation & training.
Efficient solid-waste management prototype; Top-10 at Karthavya Hackathon.
Excel-based analysis to study customers, find peak hours, and plan offers to increase sales.
Data cleaning, feature engineering, model selection & tuning (LogReg, SVM, RF, XGBoost). Clear evaluation & deployment-ready outputs.
Neural networks for NLP/CV; training & fine-tuning pipelines, metrics tracking, and export for inference.
Landing pages, portfolios, and small apps built with accessible, responsive front-end.
Exploratory analysis, visualizations, and actionable insights in Excel, Python, and R.
Clean layouts, performance tweaks, and UX improvements that make products feel great.
Send me a message—I’ll reply soon.
Don’t fill this out: