Workspace Disruptor: Former Meta AI Director Launches Slate to Replace Legacy Office Suites

After eight years leading artificial intelligence research at Meta, Rohit Patel left the social media giant to launch an AI-native replacement for Microsoft Office and Google Workspace. His new startup, named Slate, is building an intelligent workspace platform designed from scratch around direct model interactions. Patel’s high-profile departure adds to a series of recent executive exits from Meta’s AI division, following former chief medical officer Alex Peysakhovich and head of product Emily Glass-Smith.

Instead of dropping chat widgets on top of existing software, Patel designed Slate to unify document creation, data analysis, and task tracking into a single workspace. The interface features three main panels: a canvas area for drafting documents, a side panel for data tables and charts, and an active assistant panel. As users write text or update spreadsheets, the assistant continuously analyzes project context, creates data visualizations, and generates Python code in real time.

This design reflects Patel’s technical focus on reinforcement learning. During his time leading Meta’s Superintelligence Lab, Patel focused on building systems that learn through direct trial, feedback, and execution. He believes enterprise teams need software that does real work rather than just generating static text drafts. In Slate, users can inspect every auto-generated line of code, verify background calculations, and run custom Python scripts directly inside the main document view.

Patel emphasizes that Slate’s assistant maintains persistent context across an entire team’s workspace. The platform remembers project histories, internal brand guidelines, past datasets, and individual preferences. When a user pulls up a sales spreadsheet or requests a market analysis, the assistant incorporates historical company data automatically, eliminating the need to copy and paste background context into chat prompts.

Building an AI-native workspace from the ground up gives Slate significant performance and cost advantages over legacy tools. Traditional office suites rely on decades of legacy code, forcing AI add-ons to route requests through slow internal APIs. Slate processes queries directly through custom model pipelines, lowering computing overhead while speeding up real-time document updates.

Patel also pointed out that leaving Meta allowed him to move faster and make direct product decisions without navigating corporate bureaucracy. Large tech platforms often struggle to deploy aggressive product changes because every feature update requires approval from dozens of internal committees. Working with a small, focused engineering team lets Slate test new ideas, ship updates, and refine user experience patterns in response to real user feedback.

Competing against entrenched giants like Microsoft and Google remains a massive challenge. Legacy office suites dominate corporate enterprise contracts, making it hard to convince large organizations to switch primary software providers. Additionally, rivals like OpenAI and Anthropic continue to expand their own workspace tools. However, Patel argues that adding chat widgets to old software fails to solve core productivity headaches.

Slate is currently running an early preview version for select users on its waiting list, with plans to launch a public beta version in the coming weeks. By replacing fragmented office tools with an integrated, code-capable workspace, Patel aims to build a faster, smarter alternative for modern teams.