TL;DR
- Enterprise AI: Meta has started a business-focused AI platform effort and named former MongoDB chief CJ Desai to lead it for corporate customers.
- Existing Tools: The planned platform would bring Meta’s customer-chat agent, personal task assistant and developer tools to companies; each currently serves a different audience.
- Buyer Terms: Meta has not announced combined-platform access, pricing or company-wide controls, leaving business deployment terms to be defined.
Meta is starting a new enterprise AI business called Meta Enterprise Platform and has named former MongoDB chief Chirantan “CJ” Desai to lead it. As incoming Chief Enterprise Platform Officer, Desai will report directly to CEO Mark Zuckerberg as Meta works to bring its agents and developer tools into companies. Meta has not announced how business buyers would purchase or administer them as one package.
Existing Tools Serve Different Users
Meta Business Agent already handles customer conversations in WhatsApp and Messenger, answering questions, recommending products and passing harder requests to staff. In June, Meta said more than one million businesses had used it on those services and described an expansion to Instagram.
The June announcement also described a Meta Business Agent Platform for larger companies to customize those messaging agents. The similarly named Meta Enterprise Platform has a broader stated scope, including a personal agent and developer tools. Meta has not said whether the two platforms share a contract or administration system.
Muse is a personal assistant that can use a browser and connected apps to act for an individual. Meta began rolling it out in the United States on September 8.
Meta named Muse API and Muse Code for its new business. Meta’s Model API already lets developers put its AI models in their own applications, and a Meta example calls a direct request to that service a “Muse API” call. Muse Code is an existing coding agent for terminals and automated development workflows.
What Corporate Buyers Still Need
Company use would require decisions about who grants an agent access to shared data and who reviews its actions. Personal Muse asks an individual to approve sensitive actions and keeps an activity record for that user. Desai says Meta’s enterprise products build in security and privacy, but the September announcement does not define an administrator’s role for company data.
Meta’s published Model API terms give developers a price and data-use choice: the Standard tier excludes customer prompts and responses from model training, while the lower-priced Contributor tier permits training on them. The September announcement does not say whether those developer price and data rules would carry into Meta Enterprise Platform.