An expert team of engineers, AI researchers, risk quants, and builders who believe autonomous systems should be trustworthy by construction — and provable.
We started Molt because the industry’s answer to AI risk was to deploy fast and hope. We think trust should be something you build, prove, and maintain. Fisher is where we started; Deep Model Trust is where we’re going. mace AI’s work spans adversarial agent red teaming through Fisher, independent model red teaming, and agentic swarm defense to help companies prevent harm from their own multi-agent systems.
Quant Trader, Mathematician, Investor and serial entrepreneur having helped lead/participated in multiple $billion+ exits. Founder and Director of XBTO, proprietary trading firm. Graduate of Columbia University.
Previously at NVIDIA, Thomson builds the software infrastructure for secure, compliant, and accountable enterprise AI agents. His work focuses on making agents first-class participants in business operations through controlled authority, delegation, and automation. Holds an M.S. in Artificial Intelligence from Colorado State University.
Previously at Microsoft, Greg leads mace AI’s research agenda for mechanistic model innovations and architected Fisher’s evolutionary vulnerability discovery and reproducible evidence pipeline. Graduate of Harvard University.
Founder and CEO of two successful quant trading companies. Started his career trading and structuring equity derivatives on Wall Street. Brings technical and management expertise to mace AI's client execution and commercial businesses. Graduate of Macalester College, London School of Economics and Harvard University.
Previously Director of Marketing at XBTO, co-founded a successful marketing agency and a social trading platform, with earlier roles across venture capital and real estate. Brings category positioning and go-to-market expertise to mace AI's commercial growth.
Talk to mace AI about a model, configured agent, or multi-agent system.
Book demo →