Khosla Ventures is leading a $27 million (Rs 258 crore) seed round into Pramaana Labs, betting that the Bengaluru startup can make artificial intelligence more dependable in highly regulated fields. Accel, BoldCap, Nexus Venture Partners, Premji Invest, and Unbound also joined the round, backing a model that prioritizes provable correctness over probabilistic guesses.
Pramaana Labs, founded in 2025 by IIT Madras alumni Ranjan Rajagopalan, Krishnan Raghavan, and Sanjay Ganapathy Subramaniam, is building a verification layer that sits on top of AI systems. Its pitch targets sectors where factual precision is non-negotiable, including law, finance, healthcare, and taxation, and where an incorrect answer can carry legal or financial consequences.
The company is tackling hallucinations—instances where AI systems deliver confidently worded but incorrect outputs—as a first-order technical and trust problem. By treating reliability as infrastructure, Pramaana wants to ensure that AI-generated responses can be tested against authoritative rules before they are used in sensitive workflows.
Pramaana plans to deploy the newly raised capital into deepening its core technology, hiring additional researchers, and extending its capabilities across more regulated domains. The company said it is concentrating on building systems that assess whether an AI-produced answer is demonstrably accurate before it reaches professionals in law, taxation, finance, or healthcare.
Rather than functioning like conventional AI models that rely largely on statistical likelihood, Pramaana’s approach is centered on validation. Its technology cross-checks responses against codified rules and regulations, and if it cannot verify an answer, the system is designed to withhold a reply rather than risk delivering something misleading.
To make this possible, Pramaana translates complex regulatory material—such as tax codes and legal provisions—into formats that machines can systematically verify. This structure allows AI-generated outputs to be evaluated against established guidelines in real time, lowering the probability that incorrect or non-compliant information reaches end users.
The funding round lands amid sustained investor enthusiasm for AI startups worldwide and in India specifically. Interest is especially strong in companies working on foundational layers, including models, infrastructure, and trust-and-safety technologies that underpin broader enterprise adoption.
Pramaana’s raise closely follows another major deal backed by Khosla Ventures. Sarvam, an AI startup also operating out of India, recently became a unicorn after securing $300 million in a Series B round at a $1.5 billion valuation. While the companies focus on different parts of the stack, the capital flowing to both signals rising confidence in India’s ability to produce globally relevant AI infrastructure firms.
Taken together, the two financings highlight intensifying momentum behind Indian AI startups, particularly those building core technologies rather than just applications. For investors, Pramaana’s focus on verifiability addresses mounting concerns around AI hallucinations and regulatory exposure, positioning the company at the intersection of innovation, compliance, and trust.
Read Article: Perplexity Launches ‘Brain’ Self-Improving Memory System for AI Agents
