Meta just made its boldest enterprise move yet. The company pulled Dev Ittycheria, the CEO who built MongoDB into a $30 billion database powerhouse, out of his chair to run a brand new enterprise AI division. This is not about social media. This is Meta going after your company’s software budget, and it is happening right now.
What Just Happened
Meta confirmed that Dev Ittycheria, who spent over a decade as MongoDB’s CEO, will head a new enterprise AI platform initiative. Under his watch, MongoDB grew from a niche database tool into a core piece of infrastructure for thousands of companies worldwide. According to MongoDB’s own financial filings, the company crossed $1.8 billion in annual revenue before his departure. That is the kind of operator Meta wants running this division.
The enterprise AI market is getting crowded fast. Microsoft has Copilot baked into Office. Google has Gemini across Workspace. Salesforce has Einstein. Meta, until now, has been the odd one out, sitting on some of the most powerful open source AI models in the world with no serious commercial enterprise play. According to Gartner, enterprise AI software spending is projected to exceed $300 billion annually in 2026. Meta has been leaving that money on the table. Not anymore.
The Contrarian Take Nobody Is Saying
Most of the business press is framing this as Meta chasing Microsoft and Google. I think that framing is wrong.
Meta’s real edge is Llama. The company has distributed its open source Llama models to millions of developers worldwide. According to Meta’s own announcements, Llama models have been downloaded over 350 million times. That is a distribution network Microsoft and Google cannot replicate without destroying their pricing model. They need to protect margins. Meta does not.
Here is what most people miss: open source builds trust with the people who actually sign enterprise contracts. CTOs and engineering teams who have already run Llama in their own experiments are far more likely to buy a supported enterprise version than they are to adopt a new Microsoft or Google product from scratch. Meta is walking into enterprise sales conversations with a warm room. Everyone else knocks cold.
Ittycheria knows this play. MongoDB ran almost the exact same move. The database started as open source. Developers adopted it everywhere. Then MongoDB sold enterprise support, Atlas cloud hosting, and professional services on top of the free version. According to IDC, MongoDB’s enterprise license revenue grew at over 40% annually for three consecutive years during its peak growth phase. Meta is trying to run that playbook at ten times the scale.
Now think about what this means for your money. Two kinds of people will read this headline.
The first kind will see it as a tech story. They will note it, maybe share it, and go back to whatever they were doing. Their company will keep paying Microsoft $50 per seat per month for Copilot without questioning whether a Meta enterprise tool could do the same job for less.
The second kind will see a margin opportunity. If Meta enters the market with aggressive pricing, which is almost certain given that they do not need AI to be a profit center the way Microsoft does, then every company currently locked into expensive AI subscriptions has real negotiating power. Smart operators will use that power immediately.
If your business is managing AI tool costs across a team right now, keep your spending clean from day one. I use Wallester for issuing virtual business cards to team members for software subscriptions. It makes it simple to track spending per tool and cancel fast when something stops earning its keep.
What I Would Do Right Now
First, pay attention to which industry Meta targets first. If they go after financial services or healthcare, the money moves there. Those are the sectors willing to pay the highest per seat prices for enterprise AI, and Ittycheria knows that from his MongoDB days.
Second, if you run a business with more than ten employees, start auditing your current AI software costs today. When Meta launches with competitive pricing, you want to walk into that conversation knowing exactly what you are spending and what you are getting. That is how you negotiate from a position of strength instead of desperation.
Third, if you have any technical team, have them start experimenting with Llama now. Meta’s enterprise platform will almost certainly offer premium support and compliance features layered on top of what your team already knows. That head start matters when contracts hit the table.
On the people side, AI teams are expensive to build and easy to lose. When you are hiring engineers or AI specialists to work with whatever platform you choose, do not let payroll complexity slow you down. Gusto handles payroll and benefits in a way that actually works for small teams scaling fast. I have seen too many companies lose good people over administrative friction at exactly the wrong moment.
The Bottom Line
Meta just hired one of the best enterprise software operators alive to turn its AI models into a revenue machine. The companies paying attention now will have pricing power and product knowledge before the contracts get signed. The ones who wait will pay full price. Meta is not chasing Microsoft. It is building the floor under the entire market. That is a different kind of threat, and it starts today.
Frequently Asked Questions
What is Meta’s new enterprise AI platform?
Meta announced a new enterprise AI division that will sell AI products and services directly to business customers. It will be led by Dev Ittycheria, former CEO of MongoDB. The platform is expected to build on Meta’s existing Llama open source AI models with enterprise support, compliance tools, and managed cloud services on top.
Who is Dev Ittycheria and why does this hire matter?
Dev Ittycheria spent over a decade as CEO of MongoDB, growing it from an early stage database startup into a multi billion dollar enterprise software business. His track record of monetizing developer adopted open source technology is exactly the playbook Meta needs. This is not a caretaker hire. It is a signal that Meta intends to win enterprise contracts at scale.
How does Meta’s enterprise AI compare to Microsoft Copilot?
Microsoft Copilot is deeply embedded in Office 365 and Azure, which makes it easy to sell but also easy to resent. Meta’s platform will likely lead with Llama model flexibility and lower prices, since Meta does not need AI revenue to fund its core business. That structural difference could translate into real cost savings for buyers willing to switch.
Should my business wait for Meta’s platform before signing AI contracts?
If you are currently in negotiations for enterprise AI tools, Meta’s market entry gives you genuine negotiating power even if you never plan to switch. Incumbents will feel the pressure and respond with better deals. Use that information in your current conversations rather than waiting for Meta’s product to officially launch.
Is Meta’s Llama model reliable enough for enterprise use?
According to Meta, Llama models have been downloaded over 350 million times and are deployed by companies across every major industry. The open source adoption proves the technology works at scale. What the new enterprise division adds is the support, security, and compliance layer that large organizations require before they can write a purchase order.
