Microsoft Launches Frontier Company, a $2.5 Billion Unit to Deploy AI at Client Companies

Microsoft announced on July 2 the launch of a new operating unit called Microsoft Frontier Company, backed by a $2.5 billion budget and six thousand engineers and industry specialists. Their job won't be selling licenses or tools, but physically embedding within client organizations and building AI systems together that are meant to deliver measurable business results.

The forward-deployed engineering model

The practice behind the new unit is called forward-deployed engineering. Instead of selling software and leaving the customer to figure out the rest, the company sends its own engineers to design, deploy and maintain AI systems directly on the client's premises, until measurable results are achieved. The model was popularized by Palantir, which described this role as far back as its 2020 IPO prospectus, and is now being adopted one by one by the biggest players in cloud and AI.

Microsoft Frontier Company is headed by Rodrigo Kede Lima, the former president of Microsoft Asia, who brings thirty years of experience leading transformation at large organizations. The unit will bring together Microsoft's existing forward-deployed engineers, technical consultants, support teams and industry-specialized sales staff into a single team responsible for the entire deployment cycle, from design to maintenance.

First clients and partners

Among the new unit's first clients, Microsoft names London Stock Exchange Group, where AI is set to be built into the LSEG Workspace platform for financial analysts, as well as Unilever and Novo Nordisk. The company states that clients' data, intellectual property and competitive edge will not be used to train models in ways that could erode what sets them apart from competitors in their industry.

Microsoft Frontier Company is also meant to work alongside major consulting firms such as Accenture, Capgemini, EY, KPMG and PwC, rather than competing with them directly. The idea is to serve clients who need both the business strategy of consulting partners and hands-on engineering from Microsoft.

A race among big players over deployment

Microsoft's decision comes two days after Amazon pledged a billion dollars for a similar deployment unit, and follows earlier moves by OpenAI and Anthropic, which launched their own forward-deployed teams in May. The signal is clear: the AI market is shifting from a phase of experiments and demos to one of accounting for actual return on investment, and tech companies want to control that phase themselves rather than hand it off to outside integrators.

For Microsoft itself, this is also a way to defend its cloud market share against Amazon and Google, as clients increasingly ask not which AI model to choose, but who will actually help them achieve business results from deploying it. Companies with large AI budgets but no in-house engineering expertise are the natural targets for such services.

What it means for companies in Poland

For Polish businesses that use Microsoft's services, especially large organizations in finance, energy or industry, this adds a new option on the AI deployment map, alongside local integrators and global consulting firms. That said, this kind of service will likely reach the largest corporate clients first, not mid-sized companies, which still mostly rely on off-the-shelf tools like Copilot.

The forward-deployed engineering model also raises the question of vendor lock-in. A company that lets Microsoft's engineers design its AI systems from the ground up ties itself more tightly to Microsoft's cloud and models than it would by simply buying a license. That's a cost worth weighing when assessing such offers, regardless of any assurances about client data protection.

Sources: Microsoft launches its own AI deployment company with $2.5 billion commitment (techcrunch.com), Microsoft Frontier Company: AI engineering that amplifies and protects your intelligence (blogs.microsoft.com), Microsoft unveils $2.5B 'Frontier Company' to embed AI engineers inside customers (geekwire.com)