Ex-Meta AI Director Rohit Patel Launches Ndoc, an AI-Native Rival to Office and Workspace

The startup puts presentations, reports, spreadsheets and data analysis in one workspace, with a commercial launch planned for 2027.

Former Meta AI director Rohit Patel has launched Ndoc, an AI-powered productivity platform that aims to take on Microsoft Office and Google Workspace.

Patel spent nearly eight years at Meta building AI systems before leaving to start the company. Instead of adding AI features to existing apps, Ndoc is designed to bring document creation, data analysis and business communication into one integrated workspace.

One workspace, three sections

Ndoc covers presentations, reports, spreadsheets, data models and coding through three connected sections:

• Stage: presentations, dashboards, slide decks, simulations and interactive apps.
• Narrative: reports, charts, analysis and explanations of data.
• Model: the datasets and Python scripts behind the results.

The idea is to reduce hopping between tools such as Excel, Word, PowerPoint and data-analysis software. Users can upload a dataset, ask the AI assistant to analyse it, and get a narrative explaining the findings. Its pitch is to combine analysing information, developing insights and presenting them in one place.

Showing its work

Ndoc puts weight on transparency. When it generates spreadsheets, charts or other analysis, it also exposes the underlying Python code. Users can inspect the calculations or ask the assistant to explain a formula, so they can check results rather than take an AI answer on trust.

This reflects Patel’s background in AI development and evaluation at Meta. As AI plays a larger part in business reports and decisions, being able to see how an output was produced may matter more to enterprise users. Users can also add context about audience, tone and document requirements to make outputs fit their needs.

Rethinking the document file

Patel also wants to change how documents are shared. Ndoc can export work as a single .ndoc.html file that opens in an ordinary web browser. He plans to open-source the file format, which could let developers build compatible tools and reduce dependence on one vendor.

That could make it easier for teams to share work across systems. Ndoc is also developing connectors to bring in data from other applications, which should ease the move from existing tools.

A new challenger

Ndoc enters a crowded field. Microsoft and Google already have established productivity ecosystems, and AI companies are building tools that connect their models to workplace apps. Patel’s approach differs by building an AI-native workspace from the ground up rather than adding AI to traditional software.

Replacing familiar tools will be hard, though. Businesses have established workflows, document libraries and collaboration habits built around current platforms, and Ndoc will have to show enough value to make users switch.