Prentis AI Lab Seeks to Raise $100 Million at a $1 Billion Valuation

The Birth of an Ambitious AI Lab

In a move drawing significant attention from the tech world, Prentis AI Lab is in advanced discussions to secure $100 million in funding at a full $1 billion valuation. Founded in April 2025, the lab was not built from scratch alone — behind it stands serial entrepreneur Ritankar Das, alongside two heavyweights of the tech world: Reid Hoffman, co-founder of LinkedIn, and Mark Pincus, founder of Zynga.

What Exactly Does Prentis Do?

The lab's mission revolves around one core idea: automating routine office tasks through AI agents capable of independently controlling computers. The team is focused on training models that learn how office workers navigate between applications and documents while performing their daily tasks, with the goal of building agents that can complete these tasks without human intervention.

Among the key practical applications the lab is targeting:

• Automated processing and tracking of insurance claims
• Automation of customs duty drawback exceptions
• Workflow management across multiple enterprise systems

The lab has already signed contracts totaling $50 million with clients in the healthcare, manufacturing, and apparel sectors, and according to its promotional materials, expects to reach an annualized revenue run rate of approximately $75 million by the third quarter of 2025 — though these figures are performance-dependent and have not yet been realized. (TechCrunch has not independently verified these numbers.)

The Hive-32B Model: Efficiency Over Scale

Prentis is betting on its proprietary model, Hive-32B, claiming it outperforms major competitors — including OpenAI's GPT-5.4 and Anthropic's Claude Opus 4.6 — on two prominent benchmarks:

• WindowsAgentArena: measures a model's ability to complete full tasks within real Windows applications
• ScreenSpot-v2: tests a model's accuracy in identifying the correct UI elements on a computer screen

Beyond performance, the lab also highlights another competitive advantage: low cost. It claims its per-task cost is roughly one-tenth of what leading frontier model APIs charge, making it a far more economical option for companies looking to automate their day-to-day operations at scale.

The Target Market... and a Crowded Competitive Field

The founding team believes that office automation will soon surpass code generation to become the most widespread application of AI. However, this bet leads them into an arena crowded with formidable rivals: Anthropic, OpenAI, and Thinking Machines Lab — founded by Mira Murati — are all developing computer-use agents. Anthropic has gone a step further by acquiring the computer-use startup Vercept and bringing its founders on board.

The Founders: Exceptional Backgrounds

Leading the lab is Ritankar Das (age 31), who managed to graduate from UC Berkeley at eighteen with dual degrees in bioengineering and chemical biology, becoming the youngest recipient of the university's prestigious medal in over a century. He went on to complete a master's degree at Oxford, then withdrew from a PhD program at Cambridge to found Titan in 2014 — a holding company that builds and operates AI companies funded by exit revenues rather than external capital, drawing inspiration from the classic Berkshire Hathaway model.

As for Reid Hoffman, after a decade on Microsoft's board of directors, he announced a return to "founder mode" through his backing of drug-discovery company Manas AI. Similarly, Mark Pincus continues to manage his Reinvent Capital investment fund.

A Diverse Team from Elite AI Organizations

Prentis has assembled a team of more than 25 employees, many hailing from prominent institutions including OpenAI, Google DeepMind, Meta, Tencent, and Alibaba — giving the lab a research depth that belies its young age and lending greater credibility to its technical claims.

✦ بقلم فريق دروب أيديا