Pramaana Labs just landed a hefty $27 million in seed funding to make AI models do their math in front of the class.
Artificial intelligence models have become ubiquitous in fields like healthcare, finance, and transportation, but their lack of transparency and accountability is a pressing concern. That’s where Pramaana Labs comes in, with a system that compiles and verifies the decisions made by high-stakes AI.
Co-founded by Sid Seshadri and Abhinav Agrawal, Pramaana Labs aims to bridge the gap between AI’s impressive performance and its lack of explainability. The company’s product checks AI models against the rules of a domain, ensuring that the answers generated are accurate and trustworthy.
A compiler for high-stakes AI
The Pramaana Labs system is essentially a compiler for AI models, translating their complex computations into a form that can be easily understood and verified. This process, known as formal verification, involves checking an AI model’s answer against the rules of a domain. By doing so, the system helps identify potential errors or biases in the model, making it more reliable and accountable.
Traditional AI development focuses on testing models through simulation and experimentation, but these methods have limitations. With Pramaana Labs’ system, developers can ensure that their AI models are functioning correctly and making decisions based on sound mathematical principles.
What this means
The $27 million funding will help Pramaana Labs scale its technology and collaborate with industry leaders to integrate its system into various AI applications. As a result, users can expect more trustworthy and transparent AI models, particularly in critical domains like healthcare and finance. This means that life-or-death decisions, financial transactions, and other high-stakes situations will be backed by more robust and reliable AI.
