Rohit Patel, who spent nearly eight years building AI systems at Meta, has introduced Ndoc, a productivity platform that merges document writing, data analysis and presentation design into a single AI-native workspace. Rather than bolting AI features onto existing software, Patel built Ndoc from scratch to challenge the workflows long dominated by Microsoft Office and Google Workspace.
One workspace replacing several tools
Ndoc organizes work into three connected sections: Stage for presentations, dashboards and interactive apps; Narrative for written reports, charts and explanations; and Model for the raw datasets and Python scripts behind the analysis. The goal is to let someone upload a dataset, have the AI assistant analyze it, and receive both a written explanation and a visual presentation without switching between separate programs like Excel, Word and PowerPoint.
Verifiable outputs instead of black-box answers
A central feature is transparency. When Ndoc produces a chart, spreadsheet or statistical result, it also shows the Python code used to generate it, letting users inspect the logic or ask the assistant to walk through a specific formula. Patel's background evaluating AI systems at Meta appears to inform this emphasis: as companies lean more heavily on AI-generated reports for business decisions, the ability to audit how a conclusion was reached becomes a practical safeguard rather than a convenience. Users can also specify tone, intended audience and formatting requirements so outputs better match real business documents.
An open file format and new connectors
Ndoc exports completed work as a single .ndoc.html file viewable in any standard web browser. Patel intends to open-source this file format, a move that could allow outside developers to build compatible tools and reduce reliance on a single company's software. The platform is also building connectors to import data from other applications, a step meant to ease migration for teams already invested in existing tools.
Competing against entrenched habits
Ndoc arrives in a market where Microsoft and Google hold deeply established positions, and other AI companies are racing to link their models to workplace software. Patel's bet is that an AI-native design, built around transparency and integration from the start, offers a meaningful advantage over retrofitted features. Still, switching costs remain high: businesses have years of files, collaboration habits and internal workflows built around current platforms, and Ndoc will need to prove clear, sustained value before organizations consider replacing tools they already trust.