After eight years developing artificial intelligence at Meta, Rohit Patel has left the company to launch Ndoc, a startup he is pitching as an AI-native replacement for Microsoft Office and Google Workspace. Patel, most recently a director at Meta Superintelligence Labs, told TechCrunch, which exclusively reported the move on October 27, 2026, that Ndoc is now in an enterprise pilot and opening a limited alpha to wait-list customers, with a beta due in the coming weeks and a full launch planned for next year. The three-person, unfunded startup is a deliberately small bet against some of the largest software businesses in the world.
One Workspace Instead Of Scattered Tools
Patel designed Ndoc to replace what he calls the fragmented way people work: separate Word documents, PowerPoint decks, Tableau dashboards, and pasted ChatGPT output. His answer is a single workspace that colleagues can open and use even if they never signed up for the product.
Ndoc brings three components together in one layout. The Stage tab holds presentations of key findings, such as slide decks, dashboards, simulations, or mini apps. The Narrative tab carries the analysis behind those findings, with numbered citations and generated figures explaining the “why” behind the numbers. The Model tab houses the underlying datasets and Python scripts.
Checkable Work, Not A Black Box
The design reflects Patel’s background at Meta, where he focused on reinforcement learning and on evaluating whether AI output is trustworthy. If AI is going to build the spreadsheets and slide decks companies use to make decisions, he argues, someone should be able to check its work.
That is why the underlying scripts stay visible. “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. “There’s nothing hidden behind the curtain here.”
Reinventing The File Itself
Patel’s larger ambition is to reinvent the file itself. He says documents can be exported as a single “.ndoc.html” file that can be opened in any browser, a format he plans to open source so developers can build apps for it without the lock-in of Microsoft or Google.
That ambition was partly why he left Meta: at a big company, he said, agreeing such a format would require “hundreds of people and opinions.” He wanted to move at “the pace of what AI can do.”
He is the latest in a handful of high-profile departures from Meta, following chief global affairs officer Joel Kaplan, AI researcher Andrew Tulloch, and head of product for AI-for-work Emily Dalton Smith. Meta continues to pour money into AI, as in Meta’s $35 billion stock sale to fund AI expansion and Nvidia’s $3 billion investment in Meta.
A Crowded Race For Enterprise Work
Trying to become the platform where enterprise work happens is a huge bet for a three-person team with no funding. OpenAI, Anthropic, Meta, and Google have all released enterprise AI tools for Workspace and Microsoft 365, and last month OpenAI launched Space, a shared workspace inside ChatGPT that is a more direct competitor.
Patel acknowledges that Space could be a threat, but says the labs largely have the wrong idea. “Companies are basically attaching AI to spreadsheets and it fills your cells for you,” he said. “It just sort of kills my soul. 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.”
The labs have tried to bring their models to where people already work because changing habits is hard. Patel says he understands that, which is why Ndoc has connectors to import existing data. That founder-led approach echoes other recent launches, such as Entire, the AI-agent git forge launched by a former GitHub chief executive.
Patel will not raise funds until he has proof people love the product: “I didn’t want to take money until I had some proof that people love the product,” he said. Signing up is free, but customers will pay for usage and tokens.
Why It Matters
Ndoc matters because it tests whether workers will abandon the suites they already use for a single AI-native workspace, rather than accepting AI as a feature bolted onto spreadsheets and documents. If Patel is right, the value is not in filling cells faster, but in keeping the analysis, the evidence, and the underlying code together in one checkable place, so a decision can be traced back to its data.
The obstacles are just as clear. Changing workplace habits is hard, the biggest AI labs are already inside Office and Workspace, and a three-person team with no funding must prove, pilot by pilot, that people will switch. The coming beta and next year’s full launch will be the first real test.
Background
Patel spent eight years at Meta working on AI, most recently as a director in Meta Superintelligence Labs, with a focus on reinforcement learning and on whether AI output can be trusted. Ndoc, his new startup, is built on that work: a workspace meant to make AI-generated analysis verifiable, portable, and independent of any single vendor’s suite.
Source: TechCrunch’s exclusive report
0 Comments