Plurai has announced a $13 million seed round to develop its approach to improving AI agents with simulations and synthetic data.
Its platform creates a digital replica of a business’s systems for testing, while the company is also targeting smaller open-weight models tailored to each customer.
The simulation approach
Plurai founders Ilan Kadar and Elad Levi bring their experience at self-driving company Nexar to the field of artificial intelligence. The company's core idea is to create a digital replica of a company's systems. In this controlled environment, the software can be tested and trained through thousands of virtual scenarios, identifying errors before taking on work on the company's real system.
This platform works in conjunction with tools such as Claude Code, Codex, Copilot and Cursor, helping artificial intelligence models better respond to the specific needs of each client.
Efficiency and cost
Plurai focuses on creating smaller, specialized models that can deliver higher accuracy for specific enterprise tasks while reducing operational costs. Testing in partnership with Nvidia and Nebius has shown that this approach can be more cost-effective for businesses.
↳
READ ALSO
HACKS.GR / RELATED
AI Unlocked
Token's digital assistant violated security rules by design
↗
The challenges models face in complex environments are real. For example, in recent LiveBench 2.0 tests, the Claude Fable 5.2 and ChatGPT-6 Astra models showed success rates between 75% and 83%. Meanwhile, EON found that the Claude Opus 4.8 model had an accuracy of 76.8% in a simulated corporate environment, highlighting the room for improvement that Plurai seeks to address.
The company's recent $13 million funding round was led by Team8, with participation from U&I, Velocity, and Mercer.
>Stay connected with Hacks.gr
Set us as a preferred source on Google and find us on social media.
Preferred source
↗
end of article
Top comments (0)