Key Takeaways:
- Meta AI Veteran Launches Ndoc:Rohit Patel, after eight years developing AI at Meta, has departed to launch Ndoc, an AI-native platform aiming to replace the fragmented Microsoft Office and Google Workspace experience.
- Unified & Transparent Workflow:Ndoc integrates presentations, analyses, and underlying datasets/code into a single workspace, emphasizing AI output verification and a “nothing hidden” approach with explainable Python scripts.
- Bold Bet on Open-Source Reinvention:Patel seeks to reinvent the file itself with an open-source “.ndoc.html” format, challenging industry giants by prioritizing fast execution and a ground-up, AI-first design over incremental AI additions to existing tools.
A significant shift is underway in the enterprise AI landscape, marked by a high-profile departure from Meta and the launch of a bold new venture. Rohit Patel, a former director at Meta Superintelligence Labs with an eight-year tenure focused on reinforcement learning and AI trustworthiness, has left the tech giant. His new mission? To disrupt the very foundation of workplace productivity with Ndoc, a startup he envisions as an AI-native replacement for the ubiquitous Microsoft Office suite and Google Workspace.
Patel’s exit is not an isolated incident, reflecting a broader trend of talent migration from established players. He joins a growing list of high-profile departures from Meta, including Chief Global Affairs Officer Joel Kaplan, AI researcher Andrew Tulloch, and Head of Product for AI-for-Work Emily Dalton Smith, hinting at a dynamic period of innovation and re-evaluation within the AI sector.
The Fragmented Workflow’s Folly: Ndoc’s Core Premise
At the heart of Ndoc’s creation is Patel’s conviction that current productivity tools are fundamentally ill-equipped for the AI era. He argues that simply layering AI onto existing applications like Excel, Word, PowerPoint, Tableau, or even Repl.it, leads to a fragmented and inefficient workflow. The common practice of “ChatGPT copy-paste” across disparate tools underscores this inefficiency, where data, analysis, and presentation remain disconnected.
Ndoc aims to dismantle this silos-based approach by offering a singular, unified workspace. This platform is designed to be accessible to all colleagues, irrespective of whether they are Ndoc subscribers, promoting seamless collaboration from the ground up.
An Integrated Approach: Stage, Narrative, and Model
Ndoc brings three critical components into one cohesive layout, reflecting a holistic view of data-driven work:
- The Stage Tab:This is where the culmination of work is presented. Think dynamic slide decks, interactive dashboards, complex simulations, or even mini-applications that visualize key findings.
- The Narrative Tab:Here, the “why” behind the numbers comes to life. This section hosts detailed analyses, complete with numbered citations, automatically generated figures, and explanatory text that contextualizes the data.
- The Model Tab:The bedrock of trustworthiness, this tab houses the raw, underlying datasets and the Python scripts used to generate all findings and visuals.
This structure directly leverages Patel’s extensive background in reinforcement learning and his focus on evaluating AI output trustworthiness. He firmly believes that if AI is to generate the critical spreadsheets and presentations upon which business decisions are made, its work must be auditable and transparent. Users, even those without deep coding knowledge, can query the AI chat to understand script functions or verify specific formulas. “There’s nothing hidden behind the curtain here,” Patel assures, emphasizing Ndoc’s commitment to transparency.
In practical terms, a user can upload a dataset, instruct Ndoc to perform an analysis, and then automatically generate a narrative around the findings. The platform also offers advanced contextual awareness, maintaining “memory across your person, that particular document, your team, and your company,” ensuring consistency and relevance.
Furthermore, Ndoc is built and fine-tuned using a mixture of open-source models, a strategic choice that Patel says makes the platform considerably less expensive to operate than alternatives relying on frontier AI models. While he initially anticipated cost-efficiency would be a major draw, early trial users have been most captivated by the integrated experience. “What people are excited about is that they love seeing all their end-to-end in one place,” he noted.
Reinventing the File: An Open-Source Vision
Patel’s ambition extends beyond just a new suite of tools; he aims to reinvent the very concept of the digital file. Ndoc documents can be exported as a single “.ndoc.html” file, universally accessible in any web browser. His radical plan is to open-source this format, inviting other developers to build compatible applications without the proprietary lock-in characteristic of Microsoft or Google’s ecosystems.
This open-source ethos was a primary driver for his departure from Meta. Navigating the complexities of open-sourcing a format within a large corporation would have entailed “hundreds of people and opinions,” a pace incompatible with Patel’s vision. He prioritizes speed and agility, declaring, “Humans can only move at human pace, whereas…I can move at the pace of what AI can do.” This drive for rapid execution highlights the entrepreneurial spirit required to challenge entrenched incumbents.
Battling Giants: Ndoc’s Contrarian Perspective
Entering the enterprise productivity space as a lean, three-person team with no external funding is a monumental undertaking, especially when facing industry titans like OpenAI, Anthropic, Google, and Meta itself. These established AI labs have invested heavily in enterprise AI tools, often integrating their models with existing platforms like Google Workspace and Microsoft 365. OpenAI’s recent launch of “Space” within ChatGPT represents a direct move into shared workspace territory.
Patel acknowledges the competitive landscape but remains critical of the prevailing strategy. He likens the approach of simply attaching AI to existing spreadsheets to a fundamental misunderstanding of electricity’s potential. “Companies are basically attaching AI to spreadsheets and it fills your cells for you,” he critiques. “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.” His “tractor” analogy powerfully conveys his belief that a ground-up, AI-native redesign is necessary, not just incremental automation.
Changing deeply ingrained user habits is notoriously difficult, a challenge Patel understands implicitly. To ease the transition, Ndoc includes connectors for importing existing data from familiar tools. His personal benchmark for Ndoc’s quality is stringent: “My bar was that I’m not going to release anything I don’t love using…I have a graveyard of tools that did not meet that bar,” he shared, highlighting his commitment to a truly superior user experience.
Product-Led Growth: The Path Forward
Patel’s strategy for Ndoc is distinctly product-led. He intends to validate user love and traction before seeking external funding. “I didn’t want to take money until I had some proof that people love the product,” he affirmed. As of recent announcements, Ndoc is engaged in an enterprise pilot program and has opened up an alpha version to a limited waitlist. While free to sign up, customers will pay for usage and tokens, with a beta expected in the coming weeks and a full launch slated for next year.
For Patel, speed trumps perfection. His mantra is clear: “Execution is extremely cheap…so go execute on a suboptimal plan and worry about it later. By the time you make the best plan, the execution will be done by somebody else.” This philosophy underpins Ndoc’s aggressive push to redefine enterprise productivity, even against overwhelming odds.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Bottom Line
Rohit Patel’s Ndoc represents a high-stakes gamble to fundamentally reshape how businesses create, collaborate, and consume information in the age of AI. By rejecting the incremental integration of AI into legacy systems, Ndoc proposes a radical, unified, and transparent workspace built from the ground up for intelligence. Its success will hinge on its ability to overcome the inertia of established user habits and prove that an open-source, AI-native approach can deliver superior productivity and trustworthiness, challenging the very definition of enterprise software and potentially carving out a new path for innovation against the might of tech’s biggest players.
Source:{feed_title}