Funding
Plurai Raises $13 Million for Simulation Platform That Tests AI Agents Before Launch
By News Desk ·
Key takeaways
- Plurai raised $13 million for a platform that tests AI agents against thousands of synthetic, multi-turn scenarios.
- Plurai says its approach produces over 43% fewer evaluation failures than GPT-5-mini and runs at 8 times lower cost.
- Team8, U&I Ventures and Mercer Ventures participated in an approximately $10 million early-stage round dated February 1, 2026.
- Plurai released IntellAgent, an open-source framework for multi-agent simulation, and works with NVIDIA's Nemotron models and NIM software.
Plurai, a New York startup building a reliability platform for AI agents, has raised $13 million, according to Crypto Briefing. The company stress-tests agents with synthetic scenarios before they reach real users. Plurai was founded by CEO Dr. Ilan Kadar and CTO Dr. Elad Levi, both of whom hold PhDs and have backgrounds in autonomous systems.
How Plurai's Simulation Engine Works
Plurai's engine throws thousands of synthetic, multi-turn scenarios at an agent to find where it breaks before a customer does. The company also uses a process it calls "vibe-training": teams describe the rules an agent should follow in plain language, and the platform turns those descriptions into custom small language models, or SLMs. SLMs are compact AI models built for narrow jobs, and a smaller specialist trained to enforce specific policies tends to be faster and cheaper than a giant general-purpose model policing itself.
Plurai presents its performance figures as company claims rather than independent results. The startup says its approach produces over 43% fewer evaluation failures than GPT-5-mini, the OpenAI model it uses as a benchmark, and that its platform runs at 8 times lower cost. Plurai cites sub-100ms latency for real-time guardrails, fast enough for safety checks to sit inside a live conversation, and claims up to 15 times greater edge-case coverage through its simulation techniques.
Crypto Briefing's reporting points to benchmarking efforts on complex agent tasks that reference an 8% accuracy figure. Plurai has released IntellAgent, an open-source framework for multi-agent simulation aimed at evaluating conversational AI, and holds strategic partnerships with NVIDIA, working with the chipmaker's Nemotron models and NIM software.
Plurai's Earlier Funding and Product Record
Plurai's earlier financing and product history appear in company records on akta.pro. Research compiled on the company points to an approximately $10 million seed or early-stage round dated February 1, 2026, with Team8, U&I Ventures and Mercer Ventures participating.
The same records list a May 2026 product launch of aLoRA, an activated LoRA serving architecture with real-time guardrail serving that runs hundreds of guardrails on a single GPU with sub-100ms latency and 10-32x efficiency gains. An April 2026 launch, the BARRED framework, converts policy prompts into guardrails using synthetic training data and asymmetric debate without labeled data. The records also list a 2026 placement in Gartner's Market Guide for AI Evaluation & Observability Platforms and a September 2025 NVIDIA Inception membership, with IntellAgent integrating NVIDIA NIM and Nemotron reasoning models.
IntellAgent's Open-Source Release
Plurai released IntellAgent, its open-source framework for multi-agent simulation, in the 2025 to 2026 window, according to Crypto Briefing. The company's own site describes its platform for evaluating, securing and optimizing enterprise AI agents.