AI/ML Technical Lead building foundation model, scientific AI, and production AI systems from research to deployment
🧬 Foundation Models • 🔬 Pretraining Research • 🎨 Scientific AI • 🧠 Multimodal AI • ⚡ GPU Inference • 📊 Distributed ML
🥈 Kaggle March Machine Learning Mania 2026 Silver Medal
Ranked 51st out of 3,462 teams
🧬 Foundation Model Pretraining
Focused on color, spectra, materials appearance, masked spectral modeling, DeltaE evaluation, objective ablations, scaling configs, and failure analysis.
🏆 EB-1 Extraordinary Ability Recipient
📚 First-author publications
- Nature Structural & Molecular Biology
- JACS
- Journal of Molecular Biology
- PLOS ONE
💡 Inventor on Machine Learning & Computer Vision Patents
Kaggle March Machine Learning Mania 2026 — Silver Medal
| Foundation Models | Scientific AI | Pretraining Systems |
|---|---|---|
| Masked modeling | Spectral learning | Objective ablations |
| Multimodal models | Color science | Scaling experiments |
| Evaluation harnesses | Materials appearance | Metric logging |
| Agentic systems | DeltaE / CIEDE2000 | Failure analysis |
A research-oriented repo for studying foundation model pretraining loops at a tractable scale: data construction, masked modeling, training configs, evaluation harnesses, scaling studies, and failure analysis.
Tech: PyTorch, Transformers, Pretraining, Evaluation, Scaling Configs
Speech transcription + handwriting OCR + retrieval + clinical reasoning.
Tech: Whisper, OCR, RAG, FastAPI
Ensemble learning, feature engineering, pseudo-labeling, tournament prediction.
Tech: Python, XGBoost, ML
Reasoning-focused LLM workflows, LoRA fine-tuning, evaluation, and inference experimentation.
Tech: Nemotron, PyTorch, LoRA
Bioacoustic soundscape classification and environmental audio understanding.
Tech: Audio ML, PyTorch
AI / ML
PyTorch TensorFlow XGBoost Scikit-Learn
LLMs
RAG LoRA Fine-Tuning Evaluation
Infrastructure
FastAPI Docker Kubernetes Databricks Spark
Inference
CUDA TensorRT ONNX
- Foundation model pretraining
- Training research systems
- Evaluation harnesses
- Scaling and ablation studies
- Multimodal and agentic AI systems
- GPU inference optimization
Most repositories here represent research systems built to understand, evaluate, and deploy AI models.
Topics you'll find across this profile:
- Foundation Models
- Pretraining Research
- Multimodal AI
- Agentic AI
- Distributed ML
- GPU Inference
- Evaluation & Failure Analysis
The goal isn't to showcase notebooks.
Production AI systems, not just model demos.