Recently, former Meta AI Director Rohit Patel officially unveiled Ndoc, an artificial intelligence–powered productivity startup, and introduced its namesake all-in-one office platform designed to compete with Microsoft Office and Google Workspace.
Breaking away from the conventional approach of merely embedding chatbots into traditional spreadsheets, the platform constructs a workspace from the ground up, comprising three independent tabs: “Stages,” “Narrative,” and “Models.” Users can upload raw datasets via the “Models” tab to produce analytical reports that include visible Python scripts, while the integrated AI chat feature helps explain specific scripts and computational processes, ensuring business decision-makers can verify AI-generated financial data. In terms of file formats, Ndoc supports exporting documents as open-source “.ndoc.html” files that can be directly opened in a browser.
As a technical leader with eight years of experience in Meta’s reinforcement learning and AI evaluation teams, Patel opted to assemble a lean team to drive this project forward, aligning with the rapid iteration pace of modern AI tools. Facing competition from industry giants such as Microsoft, Google, OpenAI, and Anthropic, Ndoc incorporates connectors that allow data import from mainstream office applications and relies on open-source models to reduce operational costs.
Currently, the product is available in an alpha version for early testing, with a formal commercial release planned for 2027. This innovative endeavor is gradually dismantling the closed ecosystem of traditional proprietary software, offering enterprises a new solution for collaborative office work that combines transparency with openness.