SAN FRANCISCO, June 18, 2026 (GLOBE NEWSWIRE) —
Pramaana Labs, building the layer that takes AI from probably right to provably right, has raised $27 million in seed funding led by Khosla Ventures, with Accel and additional investors.
Pramaana is the first company to apply formal verification to commercial, high-stakes domains at scale. In tax, law, healthcare, finance, and government, AI still cannot stand behind its own answers. A doctor still reads the diagnosis. A lawyer still checks the brief. A CPA still signs the return. Not because the AI is too slow. Because when it is wrong in a regulated domain, it cannot be held responsible. Pramaana removes that constraint.
The system converts complex knowledge into machine-verifiable truth. First, Pramaana encodes the actual rules of a domain, the US tax code, clinical protocols, and financial regulations into a formal language that a machine can reason over with mathematical certainty. When a user asks a question, the system translates that question into a formal statement, runs it through a proof engine, and either returns a machine-checkable proof that the answer is correct or tells the user exactly which rule breaks and why. It will refuse to answer before it proves. It has never produced a confidently wrong verified answer.
“AI has an accountability gap,” said Ranjan Rajagopalan, Co-Founder and CEO of Pramaana Labs. “The world’s hardest problems are not unsolvable. They are unformalized. Every domain where being wrong can cost someone their health, money, or freedom has rules. Pramaana encodes those rules into a form that a machine can reason over with certainty. When AI can prove its answers, the human in the loop stops being a liability shield and AI becomes what it was always supposed to be: the expert.”
“Pramaana’s work is exciting because it builds on fundamental research to bring us closer to a future of seamless, trustworthy, and autonomous AI. Trustworthy AI should not require users to think like machines or become verification experts. As AI systems take on more complex tasks, trust should come from interpretable machine-checked guarantees built into the systems themselves,” said Prof. Gireeja Ranade, EECS UC Berkeley.
Every domain Pramaana formalises begins with the people who know its rules best.
Its frontier research lab brings together leading professors from IIT Delhi, IIT Madras, and UC Berkeley working on cybersecurity and drug discovery, with sponsored research at Stanford’s Centaur Lab. Its tax formalisation effort is advised by Danny Werfel, former IRS Commissioner, and built with researchers from Yale law school & Stanford. And this pattern holds across every domain: the people who built the system are the ones encoding it.
“While these are exciting times, AI-driven innovations also carry significant risk if they’re not deployed responsibly. I see Pramaana Labs as a critical missing puzzle piece that, when plugged into existing AI solutions, helps tax filers and tax professionals achieve outcomes with greater speed, accuracy, and fidelity,” said Danny.
The company’s early backers include Pushmeet Kohli, VP at Google DeepMind, and Sriram Rajamani, Corporate VP at Microsoft CoreAI, two of the world’s leading researchers in formal verification.
The founding team combines formal methods expertise with deep experience in AI systems across the world’s largest technology companies. Ranjan Rajagopalan led Google Maps Moderation, building models that kept a planetary-scale live database accurate. Krishnan Raghavan spent three years at Glean building the first version of Glean Assistant and fighting hallucinations, until he realised that solving them is a research problem, not a product problem. Sanjay Ganapathy, a former Staff Research Engineer at Google DeepMind with a decade in AI research, was a core contributor to Gemini models. He built its tool-use system and drove post-training efforts to advance frontier AI for real-world tasks. All three are IIT Madras alumni.
The funding will be used to train the formalisation and prover models, hire research engineers, and scale domain experts across regulated verticals, including tax, human diagnosis, cybersecurity and financial compliance.
About Pramaana Labs
Pramaana Labs builds AI that converts complex human knowledge into machine-verifiable truth, where every claim is grounded, traceable, and accountable. The company is the first to apply formal verification to commercial, high-stakes domains, including tax, healthcare, financial compliance, and government policy. Founded in 2025 and headquartered in Palo Alto, Pramaana is backed by Khosla Ventures, Accel, Boldcap, Nexus Venture Partners, Premji Invest and Unbound, and additional investors. Learn more at pramaanalabs.ai
