Seminar Date: Tuesday, October 6, 2026
Time: 11:00 AM PT
Location: 67-3111 & Zoom
Talk Title: AI-Ready by Design: Building DOE’s Materials Data Infrastructure from Instrument to National Platform
Zoom link
Add this event to your calendar
Abstract:
The promise of AI-accelerated materials discovery depends on a prerequisite that is rarely discussed openly: the data must actually be ready. Most experimental bench data never leaves the instrument it was collected on. Most computed data sits in siloed repositories, inaccessible to the workflows that need it. And most AI tools in materials science are starved for the structured, contextualized, FAIR ground truth they require to generalize.
This talk presents a coherent architectural response to that problem, instantiated across three complementary programs at Lawrence Berkeley National Laboratory. The Materials Project — serving 779,000 researchers across 181 countries, and the source of datasets now benchmarked by Google, Microsoft, and Meta — demonstrates what AI-ready materials data looks like at scale: curated DFT and machine-learned interatomic potential datasets, cloud-native delivery infrastructure, and an open community ecosystem built on open-source software and community-contributed data. The American Science Cloud, DOE’s Genesis Mission integration hub spanning six national laboratories, extends that model to federated scientific computing at national scale — connecting facilities, compute resources, and AI workflows under a Zero Trust identity architecture. ADEPT, a newly launched cross-divisional program at LBNL conceived to generalize this model to the experimental bench, brings instrument data into the same pipeline through structured ingestion, medallion lakehouse architecture, and automatic DOE public-access compliance — with the Molecular Foundry’s instrument fleet as a natural next target.Together these three programs form a single bench-to-AI stack — and a concrete, open-science instantiation of the data-driven and ML strategic directions DOE is actively pursuing.
Bio:
Dr. Patrick Huck is a Senior Computing Engineer (CSE5) at Lawrence Berkeley National Laboratory (LBNL) with a Ph.D. in High Energy Nuclear Physics and 16+ years of experience acting as a vital bridge between complex scientific requirements and advanced computing infrastructure. As Co-PI and Principal Technical Architect of the Materials Project, he oversees the delivery of AI-ready materials data to 779,000 researchers across 181 countries. As Lead Technical Architect of the American Science Cloud (AmSC) of the DOE’s Genesis Mission, he authored the platform’s foundational identity and architecture governance documents and leads AmSC’s Architecture Team. He is also Lead PI of ADEPT, a newly funded cross-divisional program at LBNL bridging experimental instrumentation and computational workflows between the Energy and Computational Sciences Areas.
