
I build intelligent systems that combine deep learning, production APIs, and beautiful interfaces. Passionate about AutoML, NLP, Computer Vision, and making complex models usable.
WebGL is disabled in this browser, so the portfolio is showing a styled fallback instead of a blank canvas.
Data Scientist and AI Engineer with a strong foundation in building end-to-end intelligent systems. I love the intersection of research-grade models and polished, production-ready products. Currently exploring advanced AutoML workflows, multimodal models, and 3D visualization of ML pipelines.

Python, PyTorch, TensorFlow, Next.js, FastAPI, SHAP, Docker, and AWS connected as a solar system orbit model.
Live data pipelines feeding into statistical profiling and feature engineering.
AutoML + SHAP. Every model is interpretable by default.
FastAPI → Docker → Cloud. Dashboards, alerts, and continuous evaluation.
Production AutoML platform with full EDA, model training, SHAP explainability, Optuna tuning, and one-click deployment. Live at automl-studio.netlify.app
CGPA 8.87. Applied research in NLP, Computer Vision, Cloud Architecture and Data Visualization. 50+ open source repositories.
Built reusable APIs, intelligent dashboards, model demos, and production-grade applications using modern web + AI stacks.
1,248+ contributions across 6 continents. Drag the globe to explore where the work is happening.
1,248+ contributions across India, USA, Germany, Japan, Australia, and Singapore.
WebGL is unavailable in this browser. The Copilot 3D model cannot be rendered.
Live metrics, repository status, and language breakdown synced from active code repositories.
Full-stack AutoML cockpit. Upload data, run experiments, get SHAP explanations, tune with Optuna, deploy instantly.
Context-aware medical Q&A using SBERT + BioBERT embeddings with clean conversational UI and retrieval patterns.
Explainable text classification pipeline with preprocessing, feature importance, and live prediction dashboard.
Real-time computer vision system tracking eye state and attention using OpenCV for safety applications.
Sensor-driven risk monitoring system with alerting logic, environment analytics and C++ embedded logic.
Multimodal diffusion research — realistic and consistent text-to-image generation using transformer architecture.
Fully interactive Three.js experiences. Orbit, zoom, switch modes. This is what modern portfolios can feel like.
Explore Lorenz, Rössler, and Water Lily mathematical models. Fully interactive R3F 3D trails, adjustable integration values, and zoom controls.
Cyberpunk-themed 3D chess simulator. Featuring custom geometric pieces with metallic materials, click-to-select movement, and real-time AI game replay calculations.

Visual concept depiction of the Lord of the Mysteries (LOTM) "The Fool" themed chess board, which inspired the dark cyber-emerald styling, metallic surfaces, and glowing indicators of the 3D simulation.
workflow mode selected. WebGL is disabled in this browser, so this fallback keeps the layout usable.
FULLY INTERACTIVE • POWERED BY THREE.JS + REACT THREE FIBER

IBM Skills Network

AWS Academy

IBM Skills Network

IBM Skills Network

Oracle Certified Professional

Adobe
This is a live demo — messages are simulated and not stored.