AIbaseSource publication

Former Meta AI Director Launches Ndoc: Taking On Microsoft Office and Google Workspace

Former Meta AI director Rohit Patel has officially launched the AI productivity startup Ndoc, releasing an all-in-one office platform of the same name designed to challenge Microsoft Office and Google Workspace.

Former Meta AI director Rohit Patel has officially launched the AI productivity startup Ndoc, releasing an all-in-one office platform of the same name designed to challenge Microsoft Office and Google Workspace.

The platform breaks away from the existing model of simply embedding chatbots into traditional spreadsheets, instead building from scratch a workspace with three separate tabs: "Phases," "Narratives," and "Models." Through the "Models" tab, users can upload raw datasets and generate analysis reports containing visible Python scripts, with a built-in AI chat feature to explain specific scripts and calculations, ensuring that business decision-makers can verify AI-generated financial data. As for file formats, Ndoc supports exporting documents as open-source ".ndoc.html" files that can be opened directly in a browser.

As a tech lead with 8 years of experience in reinforcement learning and AI evaluation at Meta, Patel chose to assemble a lean team to advance the project, in keeping with the iteration speed of modern AI tools. Facing competition from giants like Microsoft, Google, OpenAI, and Anthropic, Ndoc includes connectors for importing data from mainstream office applications and relies on open-source models to reduce operating costs.

The product is currently available in an alpha version for early trial use, with the official commercial version planned for release in 2027. This innovative exploration is gradually breaking the closed ecosystem of traditional proprietary software, offering a new solution for enterprise-level office collaboration that combines transparency and openness.

Original source

AIbase

Content notes

Original publication and rights belong to the source.

Machine translation · Refer to the original