Plurai raises $13M to build the simulation infrastructure for trustworthy, self-improving AI agents

• Platform now available to all developers, after it cut design partners' development cycles from months to hours, reduced costs by 97% and failures by 82%
• Round led by Team8, with backing from CyberArk founder Udi Mokady, and Google and Microsoft executives
• Joint work with NVIDIA and Nebius for a global enterprise customer outperformed OpenAI and Anthropic frontier models on accuracy at ~44x lower cost

NEW YORK, Nov. 11, 2026 /PRNewswire/ -- Plurai, the simulation infrastructure for trustworthy, self-improving AI agents, today announced $13M in seed funding and the general availability of its platform. The platform helps enterprises find and fix failures before their agents reach customers, then keeps them reliable, accurate and cost-efficient in production with small language models (SLMs) tailored to each agent's tasks. Plurai has cut design partners' agent development cycles from months to hours, reduced failures reaching real users by 82% and lowered the cost of running reliable agents by 97%. In joint work with NVIDIA and Nebius, models Plurai post-trained for a global enterprise customer outperformed OpenAI and Anthropic frontier models on accuracy at 40x and 44x lower cost, respectively.

The round was led by Team8, with participation from U&I, Velocity and Mercer, and backing from more than 20 angel investors and industry leaders, including CyberArk founder Udi Mokady. Microsoft 365 Copilot CTO Dr. Eyal Carmi has joined Plurai's advisory board.

While AI agents are easier to create than ever, making them reliable and cost-effective with real customers is far harder. VentureBeat Pulse Research recently found that half of enterprises have deployed an AI agent or LLM feature that passed internal evaluations but still caused a customer-facing failure in production. Teams spend months manually checking their agents, but they can never anticipate every scenario agents will meet in production, and once agents are live, using frontier LLMs to run and to evaluate every interaction is too slow and too expensive to scale.

From simulation to production

Plurai builds a digital twin of an agent and its world based on the customer's own knowledge base, tools and policies. It then runs thousands of realistic and adversarial users against the agent in a sandbox, covering the long tail of edge cases and exposing policy violations, hallucinations, incorrect actions, broken tool calls, failed tasks and sensitive data leaks.

The same simulation is then used to optimize the agent. Plurai generates fixes for the failures it finds and uses the simulation data to create small language models (SLMs) tailored to each agent. In production, these SLMs play two roles: they can replace frontier LLMs to run the agent, improving accuracy at a fraction of the cost, and they act as always-on evaluators and guardrails, checking every action and enforcing company policy. New failures they catch feed back into the simulation, so the agent keeps improving over time.

Previously limited to select partners and customers, Plurai is now open to every agent builder, with a free tier. Developers can start directly from the tools they already use, including Claude Code, Codex and GitHub Copilot, and teams can simulate and improve agents built on enterprise platforms such as Salesforce Agentforce.

Founders shaped by autonomous driving

Co-founders Dr. Ilan Kadar (CEO, PhD in computer science and machine learning) and Dr. Elad Levi (CTO, PhD in mathematics) have spent the past decade bringing reliable AI systems into production in mission-critical environments. At Nexar, where Kadar was VP of AI and Levi was an AI architect, they led efforts to build real-time, scalable and reliable systems for autonomous driving companies like Waymo, which rely on simulation to validate self-driving systems before they reach the road.

Their research and open-source projects are used by hundreds of thousands of developers worldwide and Plurai's technology is grounded in the company's own original research. This includes BARRED, a new method for generating synthetic training data for custom SLMs, which was accepted at ICML 2026, one of the world's leading machine learning conferences. Other research has shown how to serve hundreds of tailored SLMs in real time on a single GPU, making them economically viable at scale.

"AI systems fail when they hit challenges down the road which no one predicted," said Dr. Kadar. "Waymo built trust by testing and training across more than 20 billion simulated miles, roughly 100 times its real-world driving. Plurai brings that same approach to AI agents: continuously challenging them with the hardest situations and rarest events before they reach users, learning from failures, and improving before those failures happen in the real world."

Validation and traction: NVIDIA, Nebius and Gartner

Plurai is already working with enterprises across AI-native and highly regulated industries, including Fortune 500 and major financial companies. In addition to the work with NVIDIA and Nebius, Plurai was named a Representative Vendor in Gartner's 2026 Market Guide for AI Evaluation and Observability Platforms, a category Gartner projects 60% of software engineering teams will adopt by 2028, up from 18% in 2025. It was also featured in Gartner's "8 Emerging Practices for Testing AI Agents."

"As agents move into core enterprise workflows, the missing layer is infrastructure that builds trust in them before they reach customers and keeps them improving in production," said Aviad Harell, Managing Partner at Team8. "Simulation will become as fundamental to the agent stack as CI/CD is to software. Ilan and Elad bring deep AI expertise and over a decade of bringing AI into high-stakes production systems, and they're the team to build that layer."

Additional Quotes

"AI is probabilistic. Enterprises are not," said Udi Mokady, founder of CyberArk, chairman of CHEQ and an angel investor in Plurai. "The founders have spent more than a decade building AI for mission-critical environments like autonomous driving, where systems don't get second chances. That experience shows in how they approach agents, and it's what separates AI that can be deployed from AI that can be trusted."

"For enterprises, evaluating AI agents on real customer data is often blocked by privacy, security and compliance constraints, which makes simulation mission-critical," said Dr. Eyal Carmi, CTO of Microsoft 365 Copilot and a member of Plurai's advisory board. "What sets Plurai apart is how faithfully it recreates each company's world, from its tools and policies to the users its agents actually face. And it doesn't stop at finding failures. It turns what it learns into fixes and production safeguards, so teams can deploy agents with confidence."

"Frontier models are incredibly capable, but using them to continuously evaluate agent behavior in production can quickly become expensive and impractical," said Reah Miyara, Senior Director of Google Models at Google Cloud, former model evaluation lead at OpenAI and an angel investor in Plurai. "What I find compelling about Plurai is the idea of using smaller, tailored models to make evaluation continuous, not just something you do on a small sample of traffic."

For more information and to get started, visit: https://www.plurai.ai/

About Plurai

Plurai is the simulation infrastructure for trustworthy, self-improving AI agents. It simulates each agent's real-world environment to find failures and fix them, then post-trains the agent's underlying models on what it learns. With Plurai, agents get more reliable, more capable and cheaper to run with every cycle: simulate, fix, post-train, repeat. Founded by Dr. Ilan Kadar and Dr. Elad Levi, Plurai is backed by Team8, U&I, Velocity, Mercer, and leading industry investors. Learn more at plurai.ai.

Media Contact
[email protected]

SOURCE Plurai