By Titik Puspa 
Prentis, a nascent artificial intelligence research lab with a singular focus on revolutionizing computer use models, is reportedly in advanced discussions to secure $100 million in funding at a valuation of $1 billion. This significant capital infusion, if finalized, would underscore the immense confidence investors have in the startup’s ambitious vision to automate routine office tasks. The company, co-founded by serial entrepreneur Ritankar Das and distinguished tech titans Reid Hoffman and Mark Pincus, officially launched in April of this year, signaling a swift entry into the increasingly competitive AI landscape.
At its core, Prentis is dedicated to developing AI agents capable of learning and executing the intricate workflows that office professionals navigate daily. This involves meticulously training models to understand how individuals interact with a myriad of documents, software applications, and digital systems. The ultimate objective is to empower these AI agents to autonomously manage and complete tasks, thereby freeing up human employees from repetitive and time-consuming duties. This strategic focus positions Prentis to tackle a broad spectrum of operational inefficiencies that plague businesses across various sectors.
The practical applications of Prentis’ technology are extensive and hold the potential for substantial operational improvements. For instance, the company envisions AI agents that can expertly handle complex insurance claims, meticulously reviewing documentation, identifying discrepancies, and initiating processing without human intervention. Similarly, Prentis aims to automate the often-arduous process of customs duty refund exceptions, eliminating the need for manual paperwork retrieval and cross-referencing. These examples highlight the startup’s commitment to developing AI solutions that directly address tangible business pain points, promising significant time and cost savings for its clientele.
The traction Prentis has already achieved in its short operational history is noteworthy. According to sources familiar with the company’s discussions, Prentis has already inked contracts valued at up to $50 million with a diverse range of clients. These early adopters include a prominent healthcare management service organization, a large-scale manufacturer, and several companies operating within the goods and clothing manufacturing sectors. This early commercial success aligns with investor materials that project an estimated annualized run rate of $75 million by the third quarter of the current year. It is crucial to note, however, that these figures, as indicated in Prentis’ pitch deck, represent estimated annualized value derived from a contracted fee equivalent to 20% of realized savings, rather than recognized revenue. These projections are explicitly stated as "performance-dependent and subject to final execution," a standard disclaimer in performance-based contracts.
Prentis is making bold claims about the performance of its proprietary AI models, particularly its Hive-32B model. The company asserts that this model surpasses the capabilities of leading AI systems, including OpenAI’s GPT-5.4 and Anthropic’s Claude Opus 4.6, on two critical computer-use benchmarks. The first, WindowsAgentArena, is designed to evaluate the end-to-end task completion capabilities of AI models within real-world Windows applications. The second, ScreenSpot-v2, specifically tests a model’s proficiency in accurately identifying and interacting with the correct on-screen controls. These benchmarks are vital for assessing an AI’s practical utility in simulating human computer interaction.
The company’s strategic advantage, as outlined in its pitch deck, appears to stem from its ability to achieve superior performance with a significantly smaller and more cost-effective model. Prentis claims its approach results in approximately a tenfold reduction in cost per task compared to current frontier APIs. This economic efficiency, the company argues, makes its solution more viable and scalable for deployment across the everyday workflows of businesses. While TechCrunch has not independently verified these benchmark results, the assertion of cost-effectiveness is a compelling proposition in a market increasingly sensitive to the total cost of AI integration.
Prentis is betting heavily on the premise that the automation of everyday office tasks will soon eclipse coding as the primary use case for artificial intelligence. This is a bold prediction in a market brimming with innovation and competition. Indeed, the development of AI agents for computer use is a pursuit shared by several major players. Anthropic, OpenAI, and Mira Murati’s Thinking Machines Lab are all reportedly engaged in similar endeavors. Anthropic, in particular, has demonstrated a proactive approach to talent acquisition in this domain, having recently acquired Vercept, a Seattle-based startup specializing in AI for computer use. This acquisition saw Vercept’s founders and their expertise integrated into Anthropic, with the startup’s product being subsequently discontinued. Such moves highlight the strategic importance and competitive intensity surrounding the development of advanced AI agents.
Prentis did not respond to TechCrunch’s request for comment regarding these developments.
The leadership at Prentis is spearheaded by CEO Ritankar Das, a figure already well-established in the AI entrepreneurial ecosystem. Das is also the founder of Titan, a holding company dedicated to building and operating AI-focused businesses. At just 31 years old, Das has an impressive academic and entrepreneurial track record. He was recognized as UC Berkeley’s youngest University Medalist in over a century, graduating at the age of 18 with dual degrees in bioengineering and chemical biology. His academic journey continued with a master’s degree in biomedical engineering from Oxford.
Das’s entrepreneurial journey took a significant turn when he founded Titan in 2014, after pausing his AI PhD program at Cambridge, where he was a Gates Cambridge Scholar. Titan operates on a model that Das describes as an intentional homage to traditional holding companies like Berkshire Hathaway, emphasizing a philosophy of funding its ventures through its own successful exits rather than relying solely on external limited partners. This self-sustaining model can offer greater autonomy and long-term strategic control.
Titan’s portfolio includes several notable AI-powered ventures. Tala Health, an AI-driven virtual care provider, successfully raised a $100 million seed round last year. Forta Health, a startup focused on autism care, secured $55 million in a Series A funding round led by Insight Partners in 2024. Furthermore, Dascena, a company specializing in AI-driven disease prediction, was acquired by CirrusDx in 2022, showcasing Titan’s ability to build and exit valuable AI businesses.
The involvement of Reid Hoffman and Mark Pincus adds significant weight and strategic guidance to Prentis. For Reid Hoffman, co-founder of LinkedIn and a partner at Greylock, Prentis represents a foray back into the "founder mode" he recently expressed a desire to pursue. Last month, Hoffman announced his departure from Microsoft’s board after nearly a decade to focus on Manas AI, another AI drug-discovery startup he is backing. Hoffman’s extensive experience includes being an early investor in OpenAI and co-founding Inflection AI with Mustafa Suleyman, a venture that saw a significant portion of its team absorbed by Microsoft in 2024.
Mark Pincus, the visionary founder of Zynga, is currently at the helm of the investment firm Reinvent Capital, with Hoffman serving as a senior advisor. Pincus also recently released his memoir, "Life at the Speed of Play," further solidifying his presence in the technology and entrepreneurship spheres. Their collective experience in building, scaling, and investing in technology companies provides Prentis with invaluable strategic direction and network access.
Despite its recent launch, Prentis has already assembled a formidable team, hiring over 25 employees. This talent pool includes researchers who have previously contributed to groundbreaking work at leading AI organizations such as OpenAI, Google DeepMind, Meta, Tencent, and Alibaba. The recruitment of such high-caliber talent from established AI leaders speaks volumes about the compelling vision and potential that Prentis presents to the industry’s top minds. The company’s website serves as a testament to its ongoing recruitment and development efforts.
The competitive landscape for AI agents capable of automating computer use is indeed intensifying. As Prentis seeks to secure its substantial funding round, it operates within a dynamic environment where major tech players and innovative startups are all vying for dominance. The success of Prentis will likely hinge on its ability to not only develop technologically superior AI models but also to effectively demonstrate their tangible business value and cost-effectiveness to a broad range of enterprise clients. The company’s early traction and the caliber of its founders and investors suggest a strong foundation for navigating this challenging yet potentially transformative market.
