Meta launches an enterprise platform built around Muse
Meta is building a new business pillar around Muse, its AI agent, and has hired MongoDB's CJ Desai to run it.

Meta is going after business software. On 28 September the company announced Meta Enterprise Platform, which it calls the next major pillar of its business, built around bringing the AI stack it has been assembling for consumers into companies instead.
The initial line-up is the Muse agent, Meta Business Agent, the Muse API and Muse Code, offered to businesses and developers. Meta's stated pitch is its scale: it already serves billions of people and hundreds of millions of businesses through advertising, and it argues that owning the models, the agents and the infrastructure in one stack is something few competitors can match.
Who Meta has hired to run it
Chirantan “CJ” Desai joins Meta as Chief Enterprise Platform Officer and will report directly to Mark Zuckerberg. Desai was previously president and chief executive of MongoDB, and before that held senior roles at ServiceNow, FireEye and Cisco. Zuckerberg framed the hire in terms of “building full-stack software and delivering results in AI, infrastructure, business applications, and security”, and said Meta would be working with Desai's industry contacts.
In a statement issued alongside the announcement, Desai said the platform would turn Meta's AI stack into products businesses can deploy themselves, and repeated that security and privacy are built into Meta's enterprise products from the outset. Zuckerberg's own framing is deliberately grander: the post opens with the belief that superintelligence will create significant new opportunities for people and businesses, and positions the enterprise push as the way Meta monetises that.
What Meta is actually selling against
Enterprise AI is the most crowded part of the technology market. Microsoft sells Copilot and Azure into the same accounts, Google has Gemini in Workspace, and OpenAI, Anthropic and a long list of start-ups are chasing the same corporate budgets from the model side. Meta's advantage is that it already has a sales relationship with millions of businesses through its advertising products, and its difficulty is that being the company businesses buy ads from is not the same as being the company they trust to run their internal agents.
Our opinion
The interesting thing about this announcement is not the product list, which is a repackaging of agents Meta has already been shipping to consumers, but the admission buried in it. Meta has spent three years telling investors that superintelligence is coming; this is the first concrete plan for what happens if the ad business that funds all of it stops growing fast enough to cover the infrastructure bills. Hiring a database company's chief executive to run it is a tell about the shape of the job: MongoDB's whole business was selling to developers and enterprise architects, which is precisely the audience Meta has never had to court. That is also the gap. Meta's enterprise credibility is unproven, its previous attempt at selling workplace software, Workplace, was shut down, and the companies it now wants as customers are the same ones that have spent the last decade treating its privacy record as a risk to be managed. Desai's first year will be a test of whether that reputation is an obstacle or a footnote.