mobilezone used Microsoft Copilot Studio together with Dynamics 365 and the Power Platform to build a new customer-facing AI agent called Mia, embedded directly on mobilezone.ch.

Instead of relying on a legacy chatbot with predefined questions and limited languages, Mia now:

• Provides 24/7, multilingual, context-aware support (German, English, French, Italian).
• Acts as a virtual concierge for store locations, opening hours, subscription options, device information, and general service questions.
• Uses curated mobilezone knowledge plus real-time product and offer data from Dynamics 365 Customer Service, Dynamics 365 Finance, and Dataverse via an MCP server, so answers are accurate and up to date.
• Escalates complex or sensitive cases to human agents in Dynamics 365 Customer Service, passing full conversation context so customers don’t have to repeat themselves.

The impact is already measurable:

• Mia is part of an agent setup that handles 1,600+ conversations per month across customer and internal scenarios.
• Mia alone supports about 1,250 active chats per month, with an engagement rate of around 47% (≈2,600 total sessions).
• Language mix so far: 70% German, 20% English, 5% French, 5% Italian.
• Conversation topics: about 30% consultations for new devices or subscriptions, 25% FAQs, and 25% order status.

By surfacing relevant offers in real time and guiding customers through questions similar to an in-store consultation, Mia helps reduce avoidable contact center handovers, improves call quality, and supports faster purchasing decisions in the online shop.

In parallel with the customer agent, mobilezone built an internal IT service desk agent called Supporto using Microsoft Copilot Studio. It is deployed in Microsoft Teams and across Microsoft 365, and it reshapes how employees request IT help.

Instead of filling out rigid forms with multiple drop-downs, employees now:

• Describe issues in natural language.
• Answer follow-up questions so the agent can distinguish between incidents and service requests.
• Attach screenshots or documents directly in the chat.

Supporto then automatically creates structured tickets via API in the service desk system, and is being trained to resolve simpler issues without opening a ticket at all.

Key outcomes so far:

• Supporto handles around 350 employee chats per month with an engagement rate of about 87%.
• With Supporto as the single point of entry for IT support, average IT resolution time has been reduced by roughly 50% in 2026.
• Tickets now arrive complete and actionable, so support teams spend less time clarifying and more time solving.
• Employees experience IT support as a conversational, intuitive interaction rather than an administrative task.

Because the agent works in German, French, Italian, and English, language barriers that previously slowed down multi-level support are largely removed. mobilezone even treats Mia and Supporto as virtual colleagues with names, roles, and line managers, which helps adoption and makes AI feel like part of the everyday team.

mobilezone chose Microsoft Copilot Studio after evaluating options with a strong focus on trust, governance, and integration. Several factors were decisive:

• Data control and security: Agents are grounded in mobilezone-owned data and governed within Microsoft’s enterprise-grade security framework.
• Deep integration: Copilot Studio connects natively with Dynamics 365, Power Platform, and Dataverse, enabling agents to pull product, finance, and service data from existing systems.
• Rapid iteration: The team can iterate quickly without heavy custom development, moving beyond generic AI responses to tailored, brand-safe conversations.
• Multilingual by design: Out-of-the-box support for German, French, Italian, and English was critical for both customer and employee use cases.

With two agents live—Mia for customers and Supporto for IT—mobilezone now handles 1,600+ chats per month and has already cut IT resolution time by about 50%. Leadership sees this as a foundation for an agentic ecosystem, not a one-off project.

Next steps include:

• Extending agents into voice-based customer service.
• Adding point-of-sale personalization in stores.
• Using agents for proactive outreach around renewals and offers.
• Deeper integration with ERP and CRM systems.

As Mark Schwarz, Chief Digital Information Officer at mobilezone, puts it, the organization has moved from “almost zero” AI to treating agents as a scalable way to create new, specialized capabilities that deliver tangible value for both customers and employees.