About Xerg is building the economic control plane for AI agent fleets, from a local-first audit CLI to a hosted workspace for teams.
The name In thermodynamics, exergy is the fraction of energy in a system that is available to do useful work. The rest dissipates as waste, no matter how much raw energy you pour in.
AI agent fleets behave the same way. Part of what you spend produces working code, resolved tickets, and shipped outcomes; the rest burns off as retries, idle loops, and misrouted models. Xerg measures the useful fraction of your AI spend and helps you recover the rest.
Founders 
Founder & CEO
Jason has spent more than 15 years building software products and companies, from founding and selling a 70-person development studio in NYC to launching AI and education ventures and, most recently, Xerg. He has developed products for more than 85 companies, including Goldman Sachs and Johnson Controls, and mentors Harvard Business School MBA students in product building.
Co-founder & CTO
Charles has spent two decades taking machine learning research to market, from research scientist at Eastman Kodak to VP of Machine Learning Algorithms at the ML platform BigML to Director of Platform at Cohu. He holds a PhD in computer science and 30 AI patents, and has published peer-reviewed research on model evaluation, the discipline behind Xerg’s cost-per-outcome metrics.
