When Diakonia University of Applied Sciences (Diak) reached out for help with their AI transformation, we recognized a familiar scenario: an academic institution ready to move beyond cautious experiments with AI toward actually using it strategically. What made this partnership special was watching two very different groups (admin staff and teachers) change how they think about AI in their daily work. We worked together with Santeri Kallio on this initiative and learned more on how to support change in the university context.
Starting Where People Are, Not Where We Want Them to Be
Diak’s staff weren’t complete beginners. As one participant put it, “AI is familiar and has been in use, for example in phones, and it’s seen as a helpful support.” So our challenge was really about helping them channel their existing curiosity into use cases that would actually make their work lives easier.
By interviewing the staff and analysing feedback data, we created a roadmap that acknowledged this reality. Instead of drowning people in theoretical possibilities, we focused on quick wins that could immediately improve their daily workflows.
The Power of Differentiated Learning Paths
We split our workshops into two tracks: one for administrators and support functions and another for teachers. This made sense because these groups had completely different worries and needs.
For Administrative Staff: From Fear to Efficiency
At first, people were both excited and worried. One person admitted: “I have to verify the information myself and be familiar with the subject.” But by the end they were expressing practical ideas:
1. Automating funding channel sorting and grant application refinement
2. Creating multi-format learning materials
3. Streamlining document summarization and mind mapping
One participant’s comment really captured this change in thinking: “The most important thing I learned was that there are different AI tools for different purposes, using them together was something we all shared.“
For Teachers: Confronting the Pedagogy Challenge
Teachers had entirely different challenges: How do you maintain academic integrity while embracing all this new tech?
The breakthrough came when teachers started seeing AI as a sophisticated tutoring partner. They got excited about using AI for:
1. Creating patient practice scenarios for healthcare students
2. Developing customized exercise sets for language learning
3. Generating assessment rubrics that automatized giving feedback to students
Addressing the Elephant in the Room: Trust and Verification
We didn’t pretend AI was perfect. Instead, we made its limitations a key part of the learning experience.
“Misinformation can be mixed with correct information, making it hard to notice,” one participant pointed out. But here’s what was interesting: this awareness actually made people more confident. They went from either blindly trusting AI or being totally skeptical to using it smartly.
We saw people developing really practical strategies:
1. Always double-checking sources, especially for academic references
2. Using AI to brainstorm, then applying their own expertise to validate
3. Running things through multiple AI tools to cross-check outputs
NotebookLM: The Unexpected Star
Google’s NotebookLM became everyone’s favorite tool, especially for dealing with huge document sets and creating different learning formats. Teachers absolutely loved it for:
1. Converting dense documents into digestible summaries
2. Creating podcast-style content from written materials
3. Ideating study materials from course readings
One teacher perfectly summed up its potential: “I can use this for processing extensive materials when I want a first-phase summary. The mind map worked beautifully when I defined the sources precisely.“
Measuring Success: Beyond the Metrics
The majority of participants (85%) gave us positive feedback, which was encouraging. But the real win came from what people were saying. We watched them go from “AI might save time” to having actual, specific plans for using it.
Admin staff stopped asking “Am I allowed to do this?” and started asking “How can I do this responsibly?”
Teachers started to shift from worrying “Will students cheat?” to thinking “How can AI actually help students learn better?”
Lessons for Other Institutions
If you’re thinking about doing something similar at your organization, here’s what we learned:
1. Start with what people already know: Most staff already use AI somehow. They need help using it strategically, not starting from scratch.
2. Separate your groups: Different roles have different concerns. Address them specifically.
3. Don’t hide the problems: Talking about when AI fails actually builds more trust.
4. Keep it practical: Every exercise should connect to actual work people do every day.
A Partnership Model That Works
What made this engagement successful was the combination technical expertise, thoughtful learning design, and Splended‘s practical, hands-on approach with Diak’s openness to change.
As one participant reflected: “The most important thing was the safe and collaborative experimentation, the tips we received, and the time given to us.” This captures what we aim for in every engagement: creating space for supported exploration that leads to confident independence.
The change from AI skepticism to strategic adoption requires reimagining how work gets done. The question now is how AI will transform academic work as an inherent part bringing value rather than risks.
