I’m Hunter, an AI Research Scientist & Engineer at Roots.ai based in Jersey City, NJ.
I bridge the gap between abstract ML research and production deployment. I specialize in scaling foundation models - Large Language Models (LLMs) and Vision-Language Models (VLMs) - for production document processing systems, while conducting research in physics-informed AI for scientific simulation. I can take cutting-edge research from papers and transform it into working code for real-world systems, and I can build production ML infrastructure that advances the state of the art.
I hold a BS in Computer Science from Drexel University (2021, 4.0 GPA) and an MS in Computer Science from Harvard University (2023, 3.94 GPA), where I advanced to PhD candidacy (ABD) before transitioning to industry. At Harvard’s CSElab, I developed generative surrogate models for molecular dynamics, achieving a 7x improvement in time-series forecasting accuracy for dynamical systems.
At Roots.ai, I’ve led parameter-efficient fine-tuning (PEFT) initiatives that improved extraction accuracy from 60% to 95% on complex structured documents. I’ve engineered full-parameter distributed training pipelines for Vision-Language Models on DGX H100 clusters, managing terabyte-scale multimodal datasets with DeepSpeed ZERO. Whether I’m implementing novel architectures from recent papers, debugging distributed training failures, or designing data pipelines for scientific simulation, I operate fluently in both the research and engineering domains - taking abstract ideas to deployed systems and back again.
What You’ll Find Here
This site showcases my research and shares practical ML knowledge. You’ll find:
- Research: Publications on LLMs, document processing, scientific ML, and AI safety (including COLING 2025, EMNLP, AIES)
- Posts: Technical tutorials on foundation models, generative AI, and scientific computing
- Projects: Open-source implementations and experiments
- Notes: Reference material on datasets, methods, and tools
Let’s Connect
I’m always interested in conversations about foundation models, document processing, scientific ML, and the intersection of AI with physical sciences. Whether you’re working on production ML systems, conducting research, or exploring these topics, feel free to reach out.