How AI helped me rethink product prototyping
When Pekka Murto joined Vend’s Builder program, he was looking to move beyond casual experimentation with AI and gain a deeper understanding of how the technology could change the way products are designed and developed. What followed was not only a new workflow for prototype testing, but also a shift in how he collaborates, builds and communicates in a highly technical product environment.
Pekka Murto, a senior product designer in our Product and Tech function (PTX Core), has recently been collaborating to create more customer-centred internal tools and systems. Although he had already started exploring AI tools independently, he felt he had reached a plateau.
“I felt that I was not getting very deep in learning AI. At that point, I was chatting with LLMs on a regular basis, maybe creating a custom GPT or Gem here and there, and thinking, is this all there is to it? Luckily, that was not the case.”
His interest in automation existed long before the Builder program, though mostly through smaller personal projects.
“I did not have very deep experience with automations, but I have had a soft spot for them for sure – mainly building Shortcuts on my phone, or even a very rudimentary Python script.”
Rethinking user testing with AI-powered prototypes
During the program, Murto focused on improving one of the most important parts of product development: learning from users early. As AI-powered prototyping tools made it possible to create interactive product concepts much faster, he saw an opportunity to rethink how teams gather insights from those prototypes.
“When new AI tools, like Figma Make, became available, you were suddenly able to whip together impressive interactive code prototypes in minutes.”
Instead of relying solely on traditional one-to-one user testing, he built a lightweight analytics workflow that connects prototypes to a simple backend system, allowing teams to observe and measure user behaviour at a broader scale during the earliest product phases.
“So I thought, why not just ask the AI to build an analytics pipeline to capture a more comprehensive and nuanced picture of users in the early phases of product development?”
Bridging the gap between design and code
While the workflow itself is still new, Murto says the biggest impact has been personal. AI tools and hands-on building have helped him work more closely with the technical side of product development, despite not having a coding background.
“The impact of the overall builder program on me personally has been fundamental – although I am not fluent in the ‘material’ I am designing in (code), AI tools and building has enabled me to come a lot closer to it, both in terms of the outputs I produce and the way how I work.”
That has also strengthened collaboration within Delivery Platforms, where technical understanding plays a central role in everyday work.
“Building and AI tooling is really important for me to do better work and communicate better with other people in the area.”
Creating space for experimentation
For organisations investing in AI capabilities, Murto believes the key is giving employees time and freedom to experiment without demanding immediate results.
“Commit time and resources to it – without an immediate pressure to justify ROI. Learning to build with AI is as much about learning the tools as it is about learning about the opportunities it presents to you.”
He also believes companies need to create an environment where experimentation and failure are part of the learning process.
“And probably goes without saying, let your employees fail too. On both of these fronts, I think the builder program we have at Vend does a great job on this front.”