It may come as a surprise that Microsoft until recently used outmoded systems and processes for employee device procurement, but the company is moving quickly to modernize on the strength of centralized processes and agentic AI.
Microsoft was previously using a fragmented device procurement process that leaders acknowledge was an “industry outlier.” Imagine going into various systems, having to review a wide range of choices (50+ in Microsoft employees’ case) and decide on a device. At Microsoft’s scale, one can easily imagine the frustration and challenges, as well as the cost implications of such inefficiencies.
The company’s approach to modernizing a core business process, with plans to accelerate those plans with AI, highlight the type of practical AI use case that can deliver high-impact business and financial results.
Microsoft’s replacement system, myDevice, needed to eliminate a patchwork of systems and practices, create and enforce a consistent refresh and lifecycle model, and provide a standardized experience for employees. MyDevice supports provisioning of devices for new hires, device refreshes for existing employees, and replacement of damaged devices. Integrated with the company’s ServiceNow platform, myDevice centralizes not only ordering but also fulfillment, while providing order status visibility.
Initial Upgrade and Roadmap
With myDevice 1.0, Microsoft put in place one consistent process across geographies and divisions, with guided purchasing based on job role. Now, roughly 20,000 employees each quarter automatically get a refresh invite, assuming there is budget available for their upgrade. This takes burden off IT, administrators, and support staff, who previously managed this process manually.
Results with myDevice to date include:
- 23% reduction in primary work device spend (saving roughly $20 million annually)
- Average cost per device reduced from $1,850 to $1,670
- 50,000+ employees equipped globally in the last fiscal year
Now, Microsoft is actively enhancing myDevice with an AI-powered experience that enables one conversation with one agent that guides an employee to a device purchase.
With the Employee Device Information (EDI or Eddie) agent, employees will participate in a guided chat to explore recommended device options, compare models, and order a new device, with AI automation managing the procurement process across back-end systems.
Built with Copilot Studio, Eddie validates the employee’s eligibility, captures the required inputs, and then offers role-aligned device choices directly in chat—complete with comparison details. Once details are confirmed, the employee submits the request and receives a trackable ServiceNow reference, which reduces follow-ups and helps them stay informed throughout the process.
The agent is supported by a multi-agent architecture, with specific agents managing catalog browsing, comparison, recommendations, and ordering, controlled by an orchestration layer. Use of agentic AI brings additional opportunities including predictive refresh cycles, deeper personalization, and tighter integration with supply chain and inventory systems.
Learnings and Recommendations
Microsoft is openly sharing key learnings through this process as it moves to catch up with industry standards and leave its outlier status behind. Its recommendations include using AI to guide decisions, not just automate tasks, because employees can otherwise struggle to make decisions amid information overload.
Other recommendations include designing for an agent-first world in which employees talk to systems through AI agents for execution and predictive insights.
Many companies have better device procurement processes than Microsoft’s previous siloed approach. Yet the company’s openness on where it was behind, and its readiness to share details on building more robust procurement and supply chain processes is worth watching for those firms considering or piloting agentic AI. One lesson I’d add to those shared by Microsoft: it built a more robust process before deploying AI, then determined how and where AI could deliver additive value once the process was operating to modern standards.
More Agentic AI Insights:
- Planner Agent Pulls Together Tasks and Data To Set User Priorities
- Microsoft 365 Copilot Updates Advance Governance, Productivity App Automations
- Microsoft Advances Enterprise-Level Control for AI Agent Estates
- Copilot Studio Tools Upgraded To Bring AI Tests Into Alignment With Human Evaluations
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
Tom Smith
Analyst
Cloud Wars, Agent and Copilot
Related Posts
