Arga Labs Secures $10M to Build Training Environments for Enterprise AI Agents
aiinsider ·
Arga Labs has raised a $10 million seed round led by General Catalyst, with participation from Box Group, Emergence, Gradient, and SV Angel. The company builds training environments designed to help AI agents operate reliably within enterprise software such as Salesforce, Workday, and email systems. Unlike typical testing setups that rely on stateless API endpoints, Arga Labs creates full digital twins of enterprise programs, replicating permission systems and webhooks. CEO and co-founder Phillip Li explained that the goal is to help agents handle ambiguous, cross-system scenarios, such as recognizing when a lead in Salesforce and a separate outreach in HubSpot refer to the same company. Traditional reinforcement learning is impractical with real enterprise software because systems like Salesforce or Outlook are difficult to reset, but Arga's simulated environments allow resets and parallel training at scale. General Catalyst's Yuri Sagalov noted that much of the economic value from AI agents will come from business applications, making repeatable sandbox environments increasingly essential.
aiinsider