Plurai Lands $13M Seed to Simulate Enterprise AI Agents

Plurai, a startup founded by former Nexar engineers, raised a $13 million seed to apply the simulation playbook from self-driving cars to enterprise AI agents, TechCrunch reported.

Co-founders Ilan Kadar and Elad Levi both come from Nexar; Kadar previously collaborated with Waymo on end-to-end deep learning simulation. Team8 led the round, with U&I, Velocity and Mercer participating, plus Google DeepMind researcher Reah Miyara as an angel.

The product builds a digital twin of a customer's electronic systems, maps the possible flows an agent might take through them, and uses the resulting synthetic data to post-train small open-weight models. It plugs into Claude Code, Codex, Copilot and Cursor.

"The moment that we saw the wave of AI agents, we saw the same pattern: That it is very easy to build a demo, but [hard] to get something that will be reliable and cost effective," Kadar told TechCrunch.

In joint tests with Nvidia and Nebius, Plurai says its post-trained models matched or beat mid-tier frontier models at sharply lower cost. Kadar framed the method this way: "We can take a large language model, we can take the task, decompose it into a smaller task using the simulation, train [an open-weight] model, and using the simulation, verify that the quality is higher or comparable to the frontier model, and an order or two orders of magnitude cost reduction."

The bet rests on a wall Kadar thinks the biggest labs cannot climb: "no matter how smart OpenAI and Anthropic model will be... they will not know your business and your policy."

Plurai joins a crowded field. The AI Weekly tracker logged 572 agents stories in the last 90 days.