The entire AI industry just admitted something it's been reluctant to say out loud: software alone won't get your enterprise to the finish line.
On July 2, Microsoft launched the Microsoft Frontier Company — a $2.5 billion operating business that will embed 6,000 engineers directly inside customer organizations. Not sell them software. Not hand them documentation. Move in.
Two days before that, Amazon Web Services announced its own $1 billion forward-deployed engineering (FDE) unit. Before that, OpenAI and Anthropic launched their own joint ventures for enterprise AI services. Within a single quarter, every major AI vendor has made the same strategic pivot: from product to embedded professional services.
That's not a coincidence. That's a confession.
The deployment Problem Nobody Wanted to Talk About
Here's a number worth sitting with: according to McKinsey's State of AI 2026 report, 78% of organizations now use AI in at least one business function — up from 55% two years ago. Progress, right?
Except the same report flags a brutal follow-up: fewer than one in ten enterprises can point to a GenAI deployment that is delivering measurable, sustained business value at scale.
Seventy-eight percent started. Ten percent shipped anything real.
That gap — from pilot to production, from demo to ROI — is exactly what the FDE model is designed to close. And the fact that Microsoft, Amazon, OpenAI, and Anthropic are all making billion-dollar bets on this model at the same time tells you how serious the problem is.
I've seen this gap up close. In conversations with CIOs and VPs of Engineering at large enterprises over the past year, the pattern is consistent: teams buy the tools, build the proof of concept, show it to leadership, get budget approval — and then stall. The model works. The integration doesn't. The change management doesn't. The governance doesn't. And six months later, the POC is gathering dust while the vendor is pitching the next one.
The industry coined a name for it: POC purgatory. Microsoft just allocated $2.5 billion to get you out.
What Microsoft Frontier Company Actually Is
Microsoft's Commercial Business CEO Judson Althoff was careful to push back on the "Forward Deployed Engineer" label, calling the new unit "the largest, most capable, outcome-driven engineering organization in the industry."
The distinction matters. Traditional FDE models send engineers to build a specific system and leave. Microsoft's framing is different: a continuous loop. Engineers embedded to co-design, deploy, and continuously improve AI systems based on measurable business outcomes — not a one-time implementation.
The unit will be led by Rodrigo Kede Lima, a 30-year industry veteran who has spent the past six years at Microsoft leading enterprise-wide transformations across the Americas and Asia. This isn't a technical hire — it's a business transformation hire. The signal is clear: this is about change management as much as engineering.
Early deployments give a concrete picture of what "outcome-driven" means in practice:
London Stock Exchange Group (LSEG): Microsoft engineers embedded into LSEG's platform, helping finance professionals query complex structured and unstructured financial data in natural language. The system is iteratively refined through real-time user testing, improving model quality and scope with each cycle.
Land O'Lakes and Unilever: Both have disclosed deployments involving AI systems built to drive measurable operational outcomes — with Microsoft engineers in the room through the full lifecycle.
Novo Nordisk: Healthcare-grade AI deployment, which carries its own compliance and governance complexity.
The model extends through a partner network including Accenture, Capgemini, EY, KPMG, and PwC — all of whom are building their own FDE practices on top of Microsoft's platform. This is the scale play: Microsoft provides the AI infrastructure and core engineers; the SI partners extend the model to every market and segment globally.
The FDE Race Is Real — and Accelerating
Microsoft's announcement didn't happen in a vacuum. Let's put the timeline in order:
- May 2026: OpenAI and Anthropic both announce joint ventures for enterprise AI services, backed by outside private equity capital.
- June 30, 2026: Amazon Web Services announces a $1 billion FDE organization, explicitly embracing the model.
- July 2, 2026: Microsoft announces the Frontier Company at $2.5B, positioning it as the largest FDE effort in the industry.
In roughly 60 days, the four most consequential AI companies in enterprise technology all made the same strategic call. The message to enterprise buyers is loud and clear: if you want AI that works, you're going to need people, not just products.
For enterprise leaders, this changes the vendor selection conversation. You're no longer just buying software — you're evaluating a services relationship. Questions like "what's the API cost?" matter a lot less than "what does the engagement model look like?" and "how do you measure success?"
What This Means for Technical Leaders (CIO, CTO, VP Engineering)
If you're responsible for AI infrastructure at an enterprise, here's the honest read on what the FDE wave means for your organization:
The build-vs-buy calculus just shifted. Historically, enterprise technical teams could justify building in-house by pointing to cost control, customization, and IP protection. The FDE model changes this — you can now access deep engineering expertise embedded in your environment, with your data, building against your specific workflows. The question isn't "build vs. buy" anymore. It's "embedded vs. standalone."
Vendor lock-in risk is real — and Microsoft is trying to address it. Althoff was explicit: customer data is not used to train models. Customer IP is not commoditized. Microsoft's platform supports model diversity — OpenAI, Anthropic, open source, and Microsoft AI can all run simultaneously. For CIOs concerned about being locked into a single model vendor, this is the right answer. But read the contracts carefully. What's in the service agreement around data usage is more important than what's in the press release.
Change management is finally getting its budget. The technical leaders I've spoken with consistently identify organizational adoption — not model quality — as the biggest blocker to production AI. Microsoft's framing of the Frontier Company as an "outcome-driven engineering organization" that includes change management expertise is a meaningful signal. If the engagement model includes transformation support, not just implementation, that's materially different from what most enterprise customers have experienced.
This accelerates AI deployment timelines. For enterprises that have been stuck in POC purgatory, the FDE model offers a path out. Embedded engineers with production experience across multiple enterprise deployments can compress timelines significantly. The question is cost and prioritization — Microsoft's 6,000 engineers won't be available to everyone equally.
What This Means for Business Leaders (CFO, COO, CMO)
The business leadership view on the FDE wave is simpler, and the stakes are just as high.
AI ROI is now a services contract, not just a software license. Enterprise AI projects that include embedded engineering support have dramatically higher success rates than self-managed deployments. If your organization is investing in AI and hasn't considered the implementation services layer, you're budgeting for the easy part and ignoring the hard part. For CFOs building AI business cases, the implementation cost is not overhead — it's what separates the 10% that deliver value from the 90% that don't.
The AI vendor landscape is consolidating around outcomes. Vendors that can't prove ROI are losing ground. Microsoft's Frontier Company is explicitly structured around measurable business outcomes, with engineers held accountable to KPIs rather than deployment milestones. This is the right model. When evaluating AI vendors, ask them what success metric they're accountable to — and whether they'll put it in the contract.
Competitive differentiation is at stake. Althoff made a pointed statement in the announcement: "there is no societal permission for an AI future that eats the intelligence of the companies it's deployed inside." That's not just a legal position — it's a competitive one. Enterprises that let AI vendors train models on their proprietary workflows, customer data, and decision-making processes are giving away their competitive advantage. The vendors who can credibly protect that intelligence while accelerating deployment will win the enterprise market.
The timeline pressure is real. Amazon, OpenAI, Anthropic, and now Microsoft are all moving toward outcome-based, embedded models simultaneously. Enterprises that delay making their AI deployment choices are not standing still — they're falling behind competitors who are signing engagement agreements now. The Fortune 500 early customers already working with Microsoft's embedded engineers (LSEG, Unilever, Land O'Lakes, Novo Nordisk) are compounding their operational advantage every quarter.
Questions Enterprise Leaders Should Be Asking Right Now
The FDE race raises specific questions your organization needs to answer in the next 90 days:
1. Do we have an outcome definition? Embedded engineers can only deliver value if you've defined what value looks like. Revenue impact, cost reduction, cycle time, error rate — pick a metric and make it the engagement benchmark.
2. How do we protect our IP? Before any AI vendor embeds engineers in your environment, audit what data they can access, how model training works, and what the contract says about your proprietary information. This is legal and technical due diligence that most enterprises are still not doing rigorously.
3. Are we partner-ready? Microsoft's ecosystem includes Accenture, EY, KPMG, PwC, and Capgemini. If you have existing relationships with these firms, the Microsoft FDE model becomes accessible at a much lower entry point. Your current SI partner relationships are relevant to your AI deployment strategy in a way they weren't six months ago.
4. What's our governance framework? Embedded engineers in your environment are only as safe as your governance model allows. If you don't have AI governance — policies around model selection, data access, output validation, and human oversight — that needs to happen before or alongside any FDE engagement.
5. Who owns AI deployment in your organization? The companies succeeding with enterprise AI have a clear internal owner: typically a VP or C-level role with cross-functional authority. If AI deployment is still being managed as an IT project without business sponsorship, the FDE model will underperform. Microsoft's Frontier Company is designed for organizations that have an executive committed to transformation, not just implementation.
The Bottom Line
The launch of Microsoft Frontier Company isn't just a product announcement — it's the industry acknowledging that the biggest barrier to enterprise AI value isn't model quality. It's deployment.
The $2.5 billion investment, the 6,000 embedded engineers, the continuous improvement model — all of it is designed to close the gap between "we have AI tools" and "we have AI outcomes." Amazon, OpenAI, and Anthropic are making the same bet with similar structures.
For enterprise leaders, the practical implication is this: the era of buying AI software and figuring out deployment on your own is ending. The vendors who will win the next 18 months are the ones who can prove ROI inside your walls, with your data, measured against your KPIs.
The question isn't whether you'll need embedded AI support. The question is which vendor earns the right to show up.
Microsoft Frontier Company is available now. Early engagement information at microsoft.com/frontier-company. Amazon's FDE organization details were announced June 30, 2026. OpenAI and Anthropic enterprise ventures were announced May 2026.
Sources: Microsoft Official Blog, TechCrunch, GeekWire, CNBC, McKinsey State of AI 2026.
