Microsoft says AI projects are moving from pilots to business transformation

Microsoft says its Frontier Company has engaged in nearly 500 AI projects with more than 200 customers across 28 countries, as businesses shift their focus from AI demonstrations to measurable operational results. The company says most customers that completed initial projects have returned for additional engagements.
- Microsoft says Frontier Company has engaged in nearly 500 projects.
- The projects involve more than 200 customers across 28 countries.
- Northern Trust cut data preparation time from two and a half weeks to three days.
- KT’s platform detects network issues in two to three minutes.
- IDC says 48% of organizations are scaling or seeing consistent AI value.
Microsoft’s model focuses on workflows
Microsoft Frontier Company works through Forward Deployed Engineering teams that collaborate directly with customers on AI projects. According to Microsoft, customers including 3M, Alaska Airlines, BMW, Core42, KT, Land O’Lakes, London Stock Exchange Group, Mastercard, McKesson, Northern Trust, Novo Nordisk, Publicis Groupe, Saint-Gobain, Siemens and Unilever are using this approach.
Microsoft says the main obstacles to scaling AI are often outside the technology itself. Companies also need to identify valuable use cases, establish trust and governance, redesign operating models and help employees work differently. The company says its projects begin with business value, redesign processes rather than isolated tasks, and put security, governance and cost management in place from the start.
Customer projects span factories, finance and telecoms
At Unilever’s Pondicherry factory, a digital twin built on NVIDIA Omniverse and hosted on Azure uses live data to simulate the factory environment. Microsoft says it can help anticipate problems, monitor performance and guide operators through fixes and recommendations.
3M and Microsoft are building an AI agent-driven workflow for customer order management, including credit checks, delinquency assessments and system updates. The project includes human-in-the-loop controls and a monitoring dashboard for approvals and visibility.
At Northern Trust, an AI-powered onboarding agent integrated with Excel and SharePoint reduced data preparation time from two and a half weeks to three days, according to Microsoft. The company says testing produced 96.5% first-pass accuracy. At South Korean telecommunications provider KT, the AIO-Net platform detects network issues in two to three minutes, down from more than ten.
Microsoft also says Saint-Gobain used Microsoft Discovery to assemble 62,000 patents and 44,000 research abstracts into a knowledge base in hours. In about six weeks, the team identified hypotheses for cement additives that researchers are testing in the lab.
Scaling remains the business challenge
Microsoft cites IDC research saying 66% of organizations report substantial or transformative results from AI, while only 48% are scaling or seeing consistent value across their organizations. The research also says 71% expect AI spending to increase by at least 10% in the next year. The infographic, sponsored by Microsoft, is identified as The Real Business Value of AI, IDC #US54985026-IG, November 2026.
Microsoft says Frontier Company was designed as a flat, fast organization with one front door for customers. Rodrigo Kede Lima, President of Microsoft Frontier Company, leads its operating business of 6,000 industry and engineering experts. The company says the goal is for customers to run, govern and improve their own AI systems after the engagement, with their intelligence protected.
Microsoft: From AI ambition to business impact


