Foundation models for mass spectrometry and biochemical omics data
Matterworks applies deep learning to mass spectrometry data for drug discovery and research. The stack—Python, PyTorch, NumPy on AWS/Azure/GCP with Apache Spark—is purpose-built for scientific computing at scale. Active projects span custom neural architectures, spectral modeling, and metabolomics integration, while hiring remains tightly focused on senior ML engineers and data specialists, reflecting early-stage focus on model development over growth.
Matterworks builds Pyxis, a platform that applies foundation models to mass spectrometry and biochemical omics data. The product enables pharma and biotech teams to predict biological properties, classify disease states, and forecast R&D outcomes across drug discovery workflows. The company serves customers at scales ranging from self-service exploration to lab services and enterprise partnerships. Headquartered in Somerville, MA, the 51–200-person team is concentrated in the United States.
Python, Java, JavaScript, PyTorch, NumPy, HDF5, Zarr for modeling; AWS, Azure, GCP for infrastructure; Apache Spark for distributed data processing.
Core deep learning models, custom neural network architectures for scientific domains, a large spectral model framework, mass spectrometry data modeling, and scalable metabolomics integration.
Other companies in the same industry, closest in size