• Microsoft is investing $2.5 billion in Frontier Company for enterprise AI projects.
• The unit will help companies use multiple AI models while protecting customer IP.

Microsoft is setting up a new operating entity focused on AI implementation for enterprise customers.

The entity, called Microsoft Frontier Company, will start with $2.5 billion in funding from Microsoft. Its initial clients include Unilever and Novo Nordisk, according to the company.

How the new unit will work

Microsoft said the group will include 6,000 employees embedded with customers through a model known as forward deployed engineering. The division will bring together existing Microsoft forward deployed engineers, technical consultants, support staff, and sales employees with industry-specific experience.

Microsoft said Frontier Company will combine industry knowledge, change management, continuous improvement experience, and enterprise AI engineering expertise.

Microsoft said the group will go beyond forward deployed engineering, with 6,000 industry and engineering experts embedded with customers to help design, deploy, and improve AI systems.

Rodrigo Kede Lima, who has been leading Microsoft’s Asia business, will serve as president of Microsoft Frontier Company.

Lima has 30 years of industry experience and has spent the past six years at Microsoft, where he led enterprise transformation work across the Americas and Asia.

The new business will advise customers on AI tools from Microsoft and external providers. It will also help connect those tools with customers’ internal data, while allowing customers to retain the output of the work rather than return it to Microsoft.

Model choice and data control

Microsoft said customer data, intellectual property, and competitive information will not be used to train models in ways that reduce the customer’s control over those assets. Reuters reported that customers will keep the results of Frontier Company’s work rather than send them back to Microsoft.

Large companies are using AI systems from multiple providers instead of relying on a single model vendor, according to Reuters. These deployments can include proprietary models, open-source models, and customised systems built around company data.

The platform is designed to support models from OpenAI, Anthropic, Microsoft AI, open-source providers, and specialised models tuned for specific industries. Microsoft said the structure allows customers to use different models for different AI workloads instead of relying on one provider across all use cases.

Using a mix of AI technologies can add cost and lengthen the time needed to generate returns from AI projects. Microsoft said Frontier Company will support customers in evaluating models, integrating them into existing systems, and switching between them when required.

Microsoft said the platform will also include tools to observe, govern, manage, and secure AI systems, including the use of FinOps practices to assess returns on AI investments.

Judson Althoff, CEO of Microsoft Commercial Business, said the new company was shaped partly by Microsoft’s own experience with Copilot.

He said Microsoft had initially tied Copilot only to OpenAI models, even as other AI models advanced. “Three years ago, when we built Copilot, we made a mistake by binding it to OpenAI models only,” Althoff told Reuters. “You wanted models to amplify your intelligence and be able to have that sort of swappability for state-of-the-art and fine-tuning.”

Althoff said customers placed more value on the combination of their data and the right models than on any single model. He said businesses also needed the ability to move between models quickly.

Althoff said the FDE effort came from Microsoft’s work with customers that are still assessing how to use AI, including whether to use one model, a family of models, or a broader approach tied to existing business processes and operations.

Microsoft has a financial stake in OpenAI, the maker of ChatGPT. Earlier this year, it also added Anthropic’s models to its Copilot AI assistant.

Early customer work

Microsoft cited London Stock Exchange Group as one customer example, saying its engineers and industry teams worked with LSEG to add AI features to LSEG Workspace. Microsoft said the system helps finance professionals search across structured and unstructured financial content.

Microsoft said the LSEG system is being refined through client feedback and real-time user testing.

Microsoft also named Land O’Lakes among the customers working with its Frontier Company teams.

AI services market

Palantir has long used forward deployed engineers, while AWS and OpenAI have recently announced customer-facing AI deployment initiatives. Anthropic has also been reported to be forming a similar enterprise AI services venture.

Palantir Technologies has been using Nvidia’s open models with large customers, while Amazon Web Services has launched a $1 billion embedded-engineer unit.

OpenAI launched its OpenAI Deployment Company in May, while Anthropic was reported to be forming a similar enterprise AI services company with financial partners. AWS has also said its forward deployed engineers work directly with customer business, engineering, and security teams to build and deploy production AI systems using the customer’s data, governance, and processes.

Patrick Moorhead, CEO of analyst firm Moor Insights & Strategy, said large businesses are concerned that heavy reliance on models from companies such as Anthropic and OpenAI could allow those AI labs to build expertise in their industries. He pointed to areas such as coding and law, where model providers could gain knowledge from enterprise use cases.

Althoff credited Palantir with popularising the forward deployed engineering job title. He said Microsoft supports more models, more data connectors, and more integrations with open systems of record compared with Palantir.

Palantir’s public filing describes forward deployed engineers as staff who help customers identify new use cases, modernise data architectures, and support data-driven initiatives.

CNBC reported that Microsoft generated about $2.1 billion from enterprise and partner services in the March quarter, up 2.5% from a year earlier.

Microsoft said it has forward deployed engineering partnerships with global systems integrators including Accenture, Capgemini, EY, KPMG, and PwC. Accenture and EY had earlier announced separate plans to work with Microsoft on AI-focused forward deployed engineering programs.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

Tech Wire Asia is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.