From experiment to everyday work

The top-right is the attractive destination: a proper enterprise implementation. But all four quadrants have their place.

Quick prototypes let you test an idea against real work before investing much time or money. Nobody enjoys tedious experiments, but they are where you discover what breaks, what needs improving and what actually works. And the bottom-right can be a destination of its own: a dependable tool for one person or a small team doesn't need to become an enterprise-wide system to be worth building.

Some ideas will justify a much larger implementation. The challenge is knowing which ones. You often learn more by building a small version, testing it and improving it than by trying to plan the perfect solution on paper.

What this looks like in HR

To bring some real experience into this discussion, I invited three senior HR leaders who are already putting AI to work extensively in their organizations.

At Robit, AI supports an impressive range of activities, including analyzing employee survey responses, improving written communication, working with Excel, preparing visualizations and developing unfinished ideas into concrete plans.

Group HR Director Pia Mutanen particularly values using AI as a thinking partner. She also emphasizes questioning the results and checking their accuracy.

“With AI, the real value often comes from working through the first result. I refine the instructions, challenge the answers and verify the details until I have a result I can confidently use.”

I think that captures something important. The first result shows what might be possible. Working through the details teaches you how to make it valuable.

At One Click LCA, Director of People & Culture Irina Nordström has found a particularly valuable application in presentation design. Her team prepares everything from onboarding materials to leadership presentations, and AI has made that work considerably easier.

“Being able to generate beautiful slides with a single click so I can actually focus on the content instead of aligning text boxes is a game-changer.”

Her team can also bring older presentations into new brand templates much faster.

These are excellent examples of why I encourage experimenting with the work already in front of us.

The work that makes AI dependable

Finding a promising application is only the beginning. The next challenge is making the results dependable enough for regular use.

At VSP, Human Resources Director Pipsa Sivula uses AI in communication, reporting, leadership materials and demanding HR matters. She emphasizes providing the right context and verifying information, especially in employment-law and collective-agreement questions.

Her view of the opportunity captures something important:

“The biggest benefit is that AI frees up time from routine preparation for situations that require human judgment, interaction and decision-making.”

That is a sensible standard for many HR workflows. AI can handle substantial preparation work while experienced professionals remain responsible for the results.

And some opportunities call for much more extensive development.

At One Click LCA, Irina also leads the internal IT team. Her team is exploring automated payroll workflows across several countries, working with local payroll partners and carefully considering privacy, security and regulatory requirements.

“Payroll is an exciting opportunity for AI, but there are many pieces to get right. We're exploring where agents can reliably reduce manual work, while making sure sensitive information is protected and the processes remain transparent and compliant.”

At Robit, Pia has already built an AI agent and is exploring scheduled tasks and better-connected tools. She also sees considerable potential in an HR-system chatbot that could answer employees' everyday questions.

These three HR leaders illustrate why it's worth exploring the whole map. Some ideas become dependable tools surprisingly quickly. Others take much more work or open the door to larger opportunities. You discover which is which by trying.

For HR leaders, my advice is to make room for that work. Encourage people to experiment safely, share what they've learned and develop the ideas that show promise. Raise the standard as a workflow takes on more responsibility.

The polished implementation may be what everyone remembers. But the tedious experiments are often what made it possible.