Case study
From old-school IT support to AI-powered self-service
Service Architect, Digital Workplace, Volvo Cars. Consultant at Qestit. Gothenburg, February 2024 to present.
Snapshot
| Scope | Enterprise Digital Assist: global internal IT support and self-service for ~22,000 office workers (of ~40,000 employees), mainly in Sweden, Europe, and the US. We handle ~6,000 phone, chat, and walk-in support contacts and ~6,000 digital IT cases a month. |
|---|---|
| My role | I own key service architecture decisions and the strategy for self-service and shift-left. I influence tooling and platform decisions; no budget ownership. Team of 11, working with stakeholder groups across Volvo Cars' digital space. |
| Focus | Self-service platform and AI chatbot, knowledge and content quality, onboarding journey, user journey mapping for process improvements, making data-driven decisions |
| Tools | Service Now, Miro, Power BI, Figma, Copilot 365, Copilot Studio, Azure DevOps |
The challenge
How might we create a coherent, functional and intuitive digital ecosystem for internal users, and move away from corporate silos, bureaucratic processes, and legacy systems?
What I did
1. Led the move to AI-enabled self-service. I am spearheading a strategic initiative to replace legacy monolith systems with a distributed self-service platform, powered by AI capabilities, and built around user needs instead of technology or bureaucratic process. A Coplit Studio AI chatbot for end-user IT support is now in pilot, ahead of a wider release.
2. Made content and knowledge ready for AI. I work strategically with content quality and knowledge management, because future AI agent and LLM-driven support is only as good as the knowledge behind it. This content work runs in parallel with the chatbot pilot, so its answers are accurate before more people use it.
3. Redesigned the onboarding journey and aligned with HR/PX. I improved the onboarding user journey and lead a strategic collaboration with HR/PX towards one unified support experience for all Volvo Cars employees. I also improved the walk-in support desks: shorter waiting times, a higher customer service level, and clearer information to users.
4. Shifted mindsets and decisions toward users and data. I used visualization and storytelling, and respectfully challenged colleagues, to open new conversations about the department's future. I helped define a clear purpose, guiding principles, missions and business capabilities, and improved data reporting to support roadmap decisions.
Results
| Result | Detail |
|---|---|
| Human-assisted contacts (phone and walk-in) | ~40% lower from Q4 2024 to Q4 2026, and expected to keep falling as the AI chatbot reaches more users |
| What drove it | Onboarding redesign, the support portal and its content, password self-service and device support changes, plus a wider shift to digital channels among newer colleagues |
| My part | Led the AI chatbot initiative and led or co-led the onboarding process, the support portal and content, and the walk-in desk improvements |
| Walk-in support desks | Shorter waiting times, a better service level, clearer opening hours and information, fewer users wrongfully sent there by our phone support agents |
| AI chatbot | In pilot. Support staff testing a specialized AI agent value fast answers without interrupting colleagues all day; end-user chatbot in pilot phase |
How AI works in practice
- Use cases: the AI chatbot handles common support issues like password and access management, software and device requests, filing support incidents, and handover to live human agents when needed.
- Guardrails: users can transfer to a person for complex tasks, and manager approval stays in place for costly or sensitive requests, as before.
- Quality: knowledge content is improved in parallel (using AI workflows), so the chatbot's answers are accurate before wider release.
- Adoption: interest is high in an engineering-heavy company where earlier AI tools spread fast. The reluctant are a small minority.
Methods and tools
Methods: design leadership, stakeholder and change management, service design, process design, user journey mapping, user research, interviews, surveys, usability testing, prototyping, visualization, AI strategy, conversational design.
Tools: Service Now, Miro, Power BI, Figma, Copilot 365, Copilot Studio, Azure DevOps.
What I learned
That driving change and shifting mindsets requires communication, storytelling, empathy, and decisiveness - all of which I leverage in my daily work here.
That communicating, navigating the decision forums, internal structures, and operational processes of a large international 100-year-old corporation, is an ongoing learning journey in itself.
A lot about the disruptive possibilities, challenges and risks of emerging AI agents, LLMs and generative AI technologies. This is a field I am curiously exploring and learning more about every day.