Plurai secures $13 million seed to evaluate AI agents with synthetic data and digital twins of corporate IT environments (Tim Fernholz/TechCrunch)
agents startup
| Source: Techmeme | Original article
Plurai, a startup using synthetic data and digital twins to evaluate AI agents, has secured a $13 million seed round.
Plurai, a startup launched by former Nexar AI researchers, announced a $13 million seed round to build a simulation platform for enterprise‑grade AI agents. The company’s founders – Ilan Kadar, its chief executive, and Elad Levi, chief technology officer – are repurposing the synthetic‑data and digital‑twin techniques that enabled self‑driving cars to be tested at scale. Plurai’s infrastructure creates virtual replicas of a corporation’s IT environment, runs thousands of simulated scenarios and automatically flags where an AI agent falters. The platform then “post‑trains” the agent to correct those weaknesses before the software reaches real users.
The raise comes at a moment when AI agents are being marketed as ready for production, yet many observers note a gap between hype and actual reliability. By providing a sandbox where failures can be uncovered and remedied in a controlled setting, Plurai aims to raise the bar for trustworthy, self‑improving AI. The approach could help firms avoid costly outages, data breaches or compliance breaches that arise when agents behave unpredictably in complex, real‑world networks.
Investors appear to view the safety angle as a market opportunity, and the seed funding will be used to expand the simulation engine, onboard early corporate customers and refine the automated improvement loop. The next steps to watch include the rollout of Plurai’s platform in pilot projects, potential partnerships with large enterprises seeking to certify their AI deployments, and whether the company can attract follow‑on capital as demand for AI‑safety tooling grows. If successful, Plurai could become a critical piece of the emerging AI‑governance stack, echoing the role that simulation plays in autonomous‑vehicle certification today.
Sources
Back to AIPULSEN