BONUS Real-Time Research for Change Leaders Using AI With Ari-Pekka Skarp

AI is changing how Scrum Masters, Agile coaches, and change leaders make sense of organizations. In this BONUS episode, Ari-Pekka Skarp shares how he uses AI as a research assistant, a workspace partner, and a way to bring qualitative research into day-to-day change leadership.

"It felt like having a research assistant."

Ari-Pekka's first serious AI use did not start in his full-time change leadership role. It started with side projects, research, psychotherapy-related work, and watching his son build websites with Lovable. That curiosity became practical when Ari-Pekka began using AI to support his PhD research. Instead of spending a full day scanning articles, he could find relevant sources, generate useful summaries, and decide where to go deeper in about half an hour. The key was not that AI replaced his judgment. Because he already knew the research area, he could evaluate the quality of what the AI returned and use it as a complement to his own expertise.

"With the help of AI, I could actually make a little more of these sidetracks."

One of the biggest changes Ari-Pekka noticed was that AI made side paths cheaper to explore. In research and change work, we often ignore interesting but uncertain threads because time and attention are limited. AI gave him a way to examine adjacent ideas without losing the main direction. At the same time, he points out a real risk: AI can guide us toward something too quickly. Sometimes a slower, more intuitive decision is more aligned with the deeper research goal. For Scrum Masters and coaches, that is an important distinction: AI can widen the field of inquiry, but it should not quietly take over the choice of where to look next.

"I can create this kind of bird's eye perspective of the organization or the given project quite quickly."

Ari-Pekka describes a major shift when he started using Visual Studio Code and AI agents as a workspace for knowledge work, not only for software. In his workspaces, he organizes goals, themes, source material, Jira data, Confluence material, meeting notes, and other organizational signals so that AI can help him analyze and visualize the big picture. This gives him a practical way to represent relationships, dependencies, open threads, and organizational responses to interventions. For change leaders, the point is not the tool itself. The deeper idea is to structure the work so AI can help process the data while the human still owns the interpretation.

"I try to form a hypothesis of what I find from the data, and then I see whether those hypotheses are correct or not."

Ari-Pekka uses his qualitative research background to teach AI how to process organizational data in a consistent way. He is not just asking for a summary. He is thinking like a researcher: what is the question, what data is available, what method should be used, and what output format will make interpretation possible? Vasco connects this to the work Scrum Masters already do: hypothesis thinking, looking for evidence beyond our own biases, and using data to test what is really happening in the system. AI makes it possible to bring those research habits into day-to-day work, instead of reserving them for long academic projects.

"Now we can just use far more extensive data sets from many different sources and integrate it with the help of AI."

Ari-Pekka links AI-supported change work to PDCA: plan, do, check, act. The cycle is familiar, but the available data has changed. Teams and organizations now generate large amounts of textual, conversational, and workflow data across tools like Jira, Teams, Slack, and Confluence. AI can help integrate those sources, but Ari-Pekka warns that organizations are missing an understanding of research methodology. When the same AI tool can produce different answers from the same data, small changes in the research question, method, and prompt matter. Scrum Masters do not need to become full-time researchers, but they do need enough discipline to ask better questions and separate method from interpretation.

"With the help of AI, anybody can have a little bit of this kind of experiment in real time."

For a practical experiment, Ari-Pekka suggests starting with meeting transcripts, with explicit consent from the people involved. A Scrum Master or Agile coach can ask AI to perform a power-relation or discourse analysis of the conversation and look for patterns in who speaks, who defines the agenda, which ideas are ignored, and how decisions emerge. Organizational psychologists and researchers have studied these dynamics for years, but the work has traditionally been slow and specialized. AI makes it possible to try a lightweight version of that analysis quickly, then use the results as a prompt for reflection, not as a final judgment.

Ari-Pekka is not only a very experience Agile Coach, but he's also a Psychotherapist, and Organizational Psychologist with over 20 years of experience working with organizations. As an author of several books on topics such as Complexity, the mind, and Mindfulness, Ari-Pekka blends deep psychological insight with practical expertise to help leaders and teams navigate the evolving landscape of work.

You can link with Ari-Pekka Skarp on LinkedIn. You can read Ari-Pekka's Finnish writing at Mielen laboratorio and his English blog at Fractal Sauna. You can also find Ari-Pekka's previous Scrum Master Toolbox Podcast episodes on his guest page.