Experience

Selected work

• Large Spectral Model (LSM)Matterworks A multimodal foundation model for mass spectrometry, built for predictive biology tasks, ingesting MS1/MS2 spectra and text metadata. Self-supervised pretraining on 1M+ samples with 100M-point inputs, scaled with multi-GPU model parallelism.
• LSM-MS2Matterworks A transformer-based architecture for MS2 retrieval and generative lipid identification across 100K+ analytes, achieving state-of-the-art results on the MassSpecGym benchmark. Trained with a joint curriculum and contrastive learning strategy to separate structurally similar molecules. [arXiv]
• Generative AI for materials discoveryFord A closed-loop platform pairing GANs with latent space optimization and process-structure models to propose new microstructures for new polymers, using differential augmentation for training with limited data. [NeurIPS Workshop Spotlight]
• Geometric deep learning for atomistic simulationFord & UW-Madison Geodesic convolutions and graph neural networks over atomic structure, for both predictive and generative tasks. Applied across molecular dynamics simulations of materials and quantum mechanical (DFT) simulations of battery chemistry.

Work Experience

1. Apr 2024 – Present Matterworks Senior Machine Learning Scientist · New York, NY (Remote) Lead foundation model research across software, ML engineering, and chemistry teams, from architecture design and pretraining through integration and customer deployment. Grew the research team from 3 to 8 through Series A, designing the hiring process and reporting to the CTO.
2. Oct 2023 – Apr 2024 Matterworks Machine Learning Scientist · Boston, MA Joined as the third ML researcher post-seed, building a physics-grounded model for absolute analyte quantification that beat the industry standard by 20% across the full metabolite panel. Set company-wide QC standards and automated filtering that removed 30% of corrupted samples.
3. Jan 2022 – Jun 2023 Ford Motor Company Machine Learning Research Scientist · Detroit, MI Research spanning various domains of AI for science: generative materials discovery loops, geometric deep learning over atomic structure, and battery health monitoring. Additional work in autonomous vehicles on object detection under nighttime lens flare conditions.
4. Aug 2018 – Jun 2020 ExxonMobil Research & Engineering Process Modeling & Simulation Engineer · Houston, TX Compositional kinetic and thermal modeling to support refinery sites globally, providing process estimates, troubleshooting advice, and optimization studies. Led reactor revamp and catalyst life studies that delivered $60M in savings.

Education

1. Dec 2021 University of Wisconsin - Madison M.S. in Computer Science, Specialization in Machine Learning
2. May 2018 University of Michigan - Ann Arbor B.S.E. in Chemical Engineering, Minor in Computer Science