About
Labvana was founded on a simple observation: many research teams have extraordinary scientific talent and fragmented software foundations.
Breakthroughs depend on coordinated data, workflows that rerun, and models connected to real experimental context. When those foundations are weak, teams are asked to adopt AI on top of ground that cannot hold it.
With design partners, on their real projects, while the software is still taking shape. The science stays theirs to publish; what we learn goes back into the platform rather than into a case study.
Founding team
Labvana comes out of self-driving labs, semiconductor process development, and peer-reviewed scientific ML.
Ph.D. with 15+ years of experience in chemistry, materials science, automation, and AI-driven R&D; former STMicroelectronics principal engineer leading semiconductor process development and former LBNL researcher pioneering self-driving labs.
View the combined publication history
Ph.D. in mechanical engineering with expertise in molecular simulation, AI, and scientific computing; former Duke and NC State postdoc leading the development of ML-driven modeling tools and HPC-powered research platforms for materials and sustainability.
View the combined publication historyPrinciples
Constraints we hold to, especially when a shortcut would be faster.
Claims are backed by evidence a reviewer can follow.
A result nobody can rerun is not yet a result.
AI proposes and executes; the scientist decides.
Your data stays yours, in formats that outlive our software.
Infrastructure is judged over years, not demos.
Careers
We are small and distributed, and we work alongside the researchers we build for. If you work on scientific software, applied ML, or research operations, we would like to hear from you — including when nothing is posted.
