`

SAN FRANCISCO — Prentis, a newly launched artificial intelligence research laboratory, is currently in talks to secure $100 million in new funding at a $1 billion valuation, according to sources familiar with the matter.

Co-founded in April by serial entrepreneur Ritankar Das alongside prominent tech investors Reid Hoffman (co-founder of LinkedIn) and Mark Pincus (founder of Zynga), the stealthy startup is setting its sights on computer-use automation—training AI agents to operate computers and manage complex, multi-system office workflows autonomously.

Key Takeaways from the Pitch
Automating Back-Office Operations: Prentis is developing models designed to automate routine enterprise workflows across software applications—such as processing complex insurance claims or navigating customs duty exceptions without requiring human paperwork handling.

Proprietary Benchmark Claims: The company highlights its proprietary Hive-32B model, claiming it outperforms top-tier market rivals like OpenAI's GPT and Anthropic's Claude on computer-use benchmarks including WindowsAgentArena and ScreenSpot-v2.

Early Revenue Traction: Prentis has reportedly secured customer agreements valued up to $50 million across healthcare, manufacturing, and consumer goods sectors, projecting an annualized run rate of $75 million by Q3, subject to performance-based savings milestones.

High-Caliber Talent: The startup has already recruited over 25 researchers and engineers hailing from top AI institutions, including OpenAI, Google DeepMind, Meta, Tencent, and Alibaba.

A Crowded Arena for AI Agents
Prentis enters a fiercely competitive landscape as major tech giants and specialized startups race to build actionable AI agents capable of controlling operating systems and browser interfaces. Heavyweights such as Anthropic, OpenAI, and Mira Murati’s Thinking Machines Lab are similarly pushing hard into computer-use capabilities, betting that task automation will form the next multi-billion-dollar frontier of enterprise software.`