Engineering cutting-edge AI systems across Deep Learning, Quantum ML, and Scientific Machine Learning.
I'm a Python-first ML Engineer & Researcher specializing in GPU-accelerated pipelines, self-supervised generative modeling, hybrid quantum-classical architectures, and scientific ML for biomedical applications.
I bring together research-grade rigor with production-level engineering, building systems that are fast, reproducible, and experiment-friendly.
A high-performance cheminformatics engine for predicting molecular binding affinities using RDKit fingerprints + GPU-boosted XGBoost.
Built with a focus on reproducibility, scientific correctness, and speed, AffinityNet performs ligand standardization, feature extraction, and evaluation with near-industrial robustness — essentially a mini-drug-discovery ML pipeline.
A recursive β-VAE training framework where the model “dreams” synthetic data, filters it, and retrains on a mixed dataset to improve manifold coverage.
Lightweight yet surprisingly powerful, LUCID demonstrates how dream-filter-retrain cycles can boost reconstruction quality and stability — a concept typically seen only in advanced research papers.
A cutting-edge hybrid architecture combining GNN-style embeddings with PennyLane-powered quantum circuits for biomedical prediction tasks.
Designed around drug-aware splits to prevent leakage, NeuroQuantNet is a clean, reproducible, research-ready pipeline showcasing real-world QML viability, not toy examples.
- Deep Learning • Generative Models • Self-Supervised Learning
- Hybrid Quantum–Classical Neural Networks (QNNs, VQCs)
- Biomedical & Scientific ML (drug response, molecular fingerprints)
- GPU-Accelerated ML (CUDA, AMP, optimized data pipelines)
- Full-Stack AI Applications & Interactive ML Systems
- Quantum–Classical Hybrid Architectures
- Self-Supervised & Generative Learning
- Feature Fusion Strategies in Deep & Quantum Models
- Scientific ML for Molecules & Drug Discovery
- Graph Neural Networks for Biomedical Data
📝 Currently submitting multiple research papers to mid-tier ML & QML journals.
- Advanced Quantum Neural Network templates
- Scaling self-supervised learning for scientific datasets
- Cloud-native GPU ML deployments
