Meta is launching a new division focused on the enterprise market. Under the name Meta Enterprise Platform , the company aims to offer its AI models, agents, and infrastructure directly to organizations and developers. Former MongoDB executive Chirantan “CJ” Desai will lead the division.
This move signals Meta’s more deliberate entry into the enterprise AI software market. Until now, the company has primarily generated revenue from advertising on its consumer platforms, while its proprietary AI technology has been used for services within Facebook, Instagram, and WhatsApp, among others. With the Enterprise Platform, AI is set to become a standalone business unit.
Meta founder Mark Zuckerberg (photo) calls the platform the company’s next major pillar. The company aims to bring together various components of its existing AI stack, including the Muse agent, Meta Business Agent, Muse API, and Muse Code.
Agents for developers and customer interaction
With this move, Meta is shifting its position in the AI market. The company has long been developing models, AI infrastructure, and applications, but now wants to more explicitly package that technology into products that businesses can deploy themselves.
Muse Code, which Meta introduced in August, focuses on software development. The programming agent can distribute complex tasks across multiple sub-agents, which execute parts of the task in parallel. Meta refers to this mechanism as “agent fan-out.”
Meta Business Agent has a different target audience. This agent is designed for customer contact via WhatsApp and can, among other things, answer questions about products, make recommendations, and resolve certain technical issues. Through the Muse API, developers can also directly access the language models behind Meta’s agents.
Meta has not yet disclosed exactly what the Enterprise Platform’s commercial model will look like. Details regarding pricing and availability are also currently unavailable.
Desai takes the helm
Meta has brought in Chirantan “CJ” Desai as Chief Enterprise Platform Officer for the new division. He reports directly to Zuckerberg. Until recently, Desai was CEO and president of MongoDB. Prior to that, he was responsible for product and engineering at Cloudflare and worked for nearly eight years at ServiceNow, where he eventually became president and COO.
This background suggests that Meta not only wants to make AI technology available but is also seeking a broader position within the enterprise software market. Desai has experience in infrastructure, security, and enterprise applications, areas where Meta itself has so far been much less visible as a provider to businesses.
Isolated environment for Muse
Security and privacy must become a key part of the enterprise offering. According to SiliconANGLE, Meta is working on, among other things, Muse Confidential VM, which is set to become available later this year. According to Meta, this will isolate instances of Muse to such an extent that even Meta itself cannot access the content. A system for continuous audits is designed to allow customers to verify that this isolation actually works.
This is particularly relevant for the business deployment of AI agents. Such agents can access corporate data and other systems and perform actions autonomously. As a result, access control, data security, and oversight of what an agent is permitted to do are key requirements for large-scale deployment.
Furthermore, Enterprise Platform could eventually extend beyond software and agents. SiliconANGLE cites earlier reports from Bloomberg and CNBC stating that Meta is considering making excess capacity in its data centers available to other companies. However, Meta has not confirmed such infrastructure services as part of Enterprise Platform in its current announcement.
With Enterprise Platform, Meta is seeking to translate its substantial investments in AI into a new revenue stream outside the advertising market. How large this business can become will depend in part on the specific products Meta builds around its AI stack and the extent to which companies are willing to view Meta as a provider of mission-critical software.