AI verification startup Pramaana Labs bags $27 mn seed funding led by Khosla Ventures

Pramaana Labs, a startup that builds verification layer for AI, has raised $27 million in seed funding led by Vinod Khosla-led Khosla Ventures. Accel, Boldcap, Nexus Venture Partners, Premji Invest and Unbound also participated in the seed funding round. 

The San Francisco, California-based startup, founded by IIT Madras alumni Ranjan Rajagopalan, Krishnan Raghavan and Sanjay Ganapathy, develops technology that mathematically verifies AI-generated outputs, helping enterprises ensure that AI answers are correct. 

Pramaana, which in Sanskrit stands for “means of valid knowledge”, aims to bring certainty and verifiability to artificial intelligence systems. 

MUST READ | Who is Aman Sanger, the Indian-origin founder whose AI startup SpaceX bought for $60 billion?

While AI can generate fluent answers, those answers are not correct every time. The company believes that AI systems fail in high-stakes domains because they are optimised for plausibility rather than correctness. It aims to combine LLMs with formal logic and mathematical verification methods so that every claim can be checked against a codified set of rules. 

“We’re going to achieve that by formalising the world’s knowledge,” Ranjan Rajagopalan, co-founder and CEO of Pramaana Labs, said in a post on X. 

DON'T MISS | Jeff Bezos says AI will create labour shortages, not replace humans

“We built Pramaana to deliver a 100% trustable experience to the domains that run on certainty: AI that is probably correct, not probabilistically correct. We turn statute and regulation into machine-verifiable code, so every output ships with mathematical proof of correctness. Our mission is to make AI take ownership of its work,” Rajagopalan explained. 

Sectors such as tax, law, finance and healthcare require certainty and cannot rely on probabilistic AI systems, according to Rajagopalan. 

“The foundational domains that hold the world together: tax, law, finance, healthcare; all run on certainty. Probabilistic AI can't give them that. We’ve been asked to accept wrong answers with AI as ‘hallucinations’, while in traditional software terms, it’s just a bug. And a wrong answer in such mission-critical domains is more than just a bug, it's a liability that could have catastrophic impact,” he said. 

MUST READ | India’s AI dream runs on a stack it does not control