Rohit Patel, who spent eight years building AI systems at Meta, has left the company to launch Ndoc, a startup positioning itself as an AI-native alternative to Microsoft Office and Google Workspace, TechCrunch has exclusively learned. The former director at Meta Superintelligence Labs joins a string of recent high-profile exits from the company, including Chief Global Affairs Officer Joel Kaplan, AI researcher Andrew Tulloch, and Head of Product for AI-for-Work Emily Dalton Smith. Patel's departure reflects a broader trend of senior talent leaving established tech giants to build specialized AI tools that challenge legacy productivity platforms.

Rather than layering AI features onto existing workflows, Ndoc was designed to consolidate the fragmented way people toggle between tools like Excel, Word, PowerPoint, Tableau, Replit, and "ChatGPT copy-paste" into a single unified workspace that colleagues can access even without signing up for the platform. The product brings together three elements in one interface: a Stage tab for presentations, slide decks, dashboards, simulations, or mini apps; a Narrative tab containing analyses, numbered citations, generated figures, and the reasoning behind the data; and a Model tab housing the underlying datasets and Python scripts. Users can upload a dataset and request that Ndoc perform an analysis and craft a narrative about the results, while also adding document-specific context including audience, tone, and memory that spans individual users, particular documents, teams, and entire companies. Patel says the platform runs on a mixture of open source models and costs considerably less to operate than alternatives built on frontier AI models, though he notes that early trial users care less about cost efficiency and more about seeing their entire workflow consolidated in one place.

Patel's design philosophy stems directly from his time at Meta, where he concentrated on reinforcement learning—a machine learning approach in which AI agents learn through trial and error to maximize rewards—and assessing the trustworthiness of AI output. "Even if you don't know any Python, you can ask [the AI chat] to explain, 'Hey, what does this script do? Which line has my formula?' And you can verify that formula," Patel said. His broader goal is to reinvent the file itself: documents can be exported as a single ".ndoc.html" file that opens in any browser, a format he plans to open-source so other developers can create compatible apps without vendor lock-in. Getting hundreds of people at a large company like Meta to agree on an open source format would've slowed the project down, which is partly why he struck out on his own. "Humans can only move at human pace, whereas…I can move at the pace of what AI can do," he explained.

Patel faces intense competition from well-funded AI labs that are also vying to control the workspace layer: OpenAI, Anthropic, Meta, SpaceXAI, and Google have all released enterprise AI tools connecting to Google Workspace, Microsoft 365, and other workplace apps, while OpenAI launched Space last month—a shared workspace inside ChatGPT that directly competes with established productivity suites. Patel acknowledges Space as a potential threat but argues the major labs have the wrong approach. "Companies are basically attaching AI to spreadsheets and it fills your cells for you," he said. "This is like if you just figured out electricity, and then you decided the best use case is to put lights on my bulls that till my soil. No. Get a tractor." As of Tuesday, Ndoc is running an enterprise pilot and opening an alpha version to a limited number of customers from its wait list—free to sign up but paid on usage and tokens—with a beta planned for the coming weeks and a full launch next year. Patel is operating with a team of three and no outside funding, waiting for proof that customers love the product before raising capital. "By the time you make the best plan, the execution will be done by somebody else," according to the report. The bet is that moving quickly with a suboptimal plan beats perfecting strategy while competitors execute. Patel's willingness to challenge entrenched productivity giants with a skeleton crew signals either visionary confidence or a fundamental misread of how deeply enterprises resist changing established workflows. Whether open-sourcing file formats can break decades of vendor lock-in remains the core question that will determine if Ndoc becomes a tractor or just another graveyard tool.