Editorial illustration for Hoffman, Pincus AI lab Prentis in talks to raise USD 100 million
Hoffman, Pincus AI lab Prentis in talks to raise USD 100...
Prentis, an AI research lab launched in April, is in talks to raise $100 million at a $1 billion valuation, according to two people familiar with the discussions. The startup was co-founded by Ritankar Das along with Reid Hoffman and Marc Pincus, two names well known in Silicon Valley for backing companies early and often. Prentis is building models that watch how office workers move through routine tasks, filling out forms, switching between documents and systems, with the aim of training AI agents that can eventually take over those tasks by controlling a computer directly.
The company has already signed contracts worth up to $50 million with a handful of customers, including a healthcare management service organization, a manufacturer, and a clothing goods maker, the two people told TechCrunch. Investor materials reviewed by TechCrunch point to an estimated $75 million annualized run rate by the third quarter of this year, though Prentis's own pitch deck cautions those numbers reflect projected value tied to a fee structure, not booked revenue, and remain "performance-dependent and subject to final execution."
Launched in April, Prentis is training models to learn how office workers navigate routine workflows across documents and systems, with the goal of building AI agents that can control computers to automate those tasks.
Why this matters
Prentis is chasing the same computer-use problem that OpenAI, Anthropic and Adept have all poured resources into: teaching models to actually operate software the way a human employee does, clicking through documents and systems rather than just answering prompts. A $1 billion valuation for a lab founded in April, with no public product yet, tells us investors are still pricing agentic AI on the promise of enterprise workflow automation rather than shipped results. That's worth watching closely if you're building in this space, because the bar for "computer use" demos has been low so far, reliability over long multi-step tasks remains the hard unsolved part.
Das's Titan structure is the more unusual detail here. Funding new ventures like Prentis and Tala Health off exits instead of LP capital is a bet that founders with cash from prior wins can outrun traditional VC timelines. For researchers and founders, the real signal to track isn't the valuation, it's whether Prentis can show office-workflow agents that hold up outside a demo reel.
Further Reading
- Papers with Code - Latest NLP Research - Papers with Code
- Hugging Face Daily Papers - Hugging Face
- ArXiv CS.CL (Computation and Language) - ArXiv