Inside Sitedrive’s product team, and the shift that’s changing how we work.
At Sitedrive, we build a production management platform that helps construction teams run projects with factory-like flow, real-time transparency, and fewer surprises. But behind the product is a small, experienced team working in a market where speed matters: construction doesn’t wait for long development cycles.
Over the past year, AI has changed how our product team thinks, collaborates, and ships. Not because it’s trendy, but because it makes us better at solving real customer problems.
The shift: From “AI as a tool” to “AI as part of the workflow”
When we asked our team what has changed most in the past 12 months, the answer was unanimous: the speed of change itself.
“We test new tools constantly. Some die out fast, some stick. The pace is just crazy,” says Lead Data Scientist Jussi Engblom. “Useful things survive, the rest falls away.”
What surprised us wasn’t only the new tools, but how naturally they became part of everyday work. We don’t “use AI” like a separate phase. It sits inside our workflows: writing code, validating ideas, clarifying logic, structuring components, documenting decisions.
Our Head of Product, Henri Ahoste describes it simply: “Every single person in our product development team uses AI now.”
Designers code, engineers prototype, everyone moves faster
One of the clearest changes AI brought to Sitedrive is role fluidity, the ability to step outside your formal specialty and contribute where it matters.
Our Lead Designer Mika Käki explains how AI reopened a door he thought had closed years ago:
“I’ve been a developer back in 2006, but haven’t really coded since. Without AI, I would have needed six months of immersion to even start. Now I can actually write production-adjacent code, all the writing happens with AI.”
This shift has two big effects:
1. Designers can directly improve user experience
Mika tells a small but meaningful example: reducing a 1-second delay in the UI.
“When I can fix that myself, I know the improvement is real for the user. I don’t need to ask a developer to polish something that might seem small. It’s empowering.”
2. Product management can ship prototypes, not just specs
Henri now uses AI to turn ideas into interactive prototypes inside the real product environment, not static mockups.
“I can visualize an idea directly in the product. It’s a working suggestion, not just a wireframe.”
3. Backenders write frontend, frontenders write backend
AI evens out the playing field. Jussi (backend) puts it bluntly:
“Front-end code is a bit tedious, but when the bot writes it, I can read and understand it easily.”
This doesn’t replace expertise, and the team is clear about that.
“Work outside your core area rarely meets our production quality bar. You still need teamwork,” Henri notes. “AI widens your capability, but your strengths still matter.”
Testing is no longer expensive. Creativity is no longer slow.
One of the biggest constraints in product development – especially for a small team – is the cost of testing ideas. Prototypes take time. Wrong ideas can be expensive.
AI flipped that equation.
“We can now test and validate multiple prototypes and throw away the ones that don’t work. That’s something only big teams could do earlier.” – Henri
Prototypes have become a core design tool:
- Mika builds functional, interactive prototypes using Google AI Studio.
- Henri built a massive prototype in weeks, one that helped shape real product direction.
- Those prototypes can be moved into our actual codebase and refined into production.
This makes design more concrete, faster:
“AI doesn’t make me more creative, but it makes my ideas visible faster. When something becomes concrete, you immediately get a better understanding of how it might work, and you can test it, too.” – Mika
Our everyday AI toolbox
We’re not precious about tools. We use what works.
Our current core stack includes:
- Claude Code – the main engine behind most code generation (~90% of outputs today)
- Google AI Studio – fast prototyping for design and interaction concepts
- Claude and Gemini – for documentation, explanation, restructuring, data thinking
- Internal agent workflows – for repetitive engineering tasks, pattern-based code, refactoring, creating and maintaining end-to-end tests.
But more important than the tools is how we work with them.
What surprised us the most
1. AI produces 10K lines of code in 5 minutes…
…and creates a new problem:
reading and validating 10,000 lines of code.
“The challenge moved. It’s no longer writing code – it’s reading, structuring, and keeping it under control.” – Henri
2. AI still lies. Boldly.
Mika laughs: “Sometimes it just confidently says something wrong. You think: what the hell is this?”
3. Good engineering practices matter more than ever
Jussi: “If you want AI to work well, your code must be clean and modular. That actually makes the whole team better.”
Top 3 lessons we’d give any SaaS team adopting AI
- Make your codebase clean and modular. AI multiplies chaos if your architecture is chaotic.
- Prototype relentlessly, and throw things away fast. AI makes it cheap, so use that freedom.
- Stay in the driver’s seat. As Henri says: “AI doesn’t know what it doesn’t know, especially your users.”
So… what does this mean for Sitedrive customers?
Short answer: faster value. Sitedrive’s product philosophy is rooted in one goal:
shorter time-to-value for construction teams.
AI helps us deliver that in several ways:
- Features validated earlier and more accurately
- More iterations in less time
- Better usability through designer-led micro-improvements
- Faster turnaround for real customer needs
- Fewer blind spots in decision-making
“Time to market gets shorter. That’s the real reason we do this,” Jussi says.
We don’t add AI to our product because it’s trendy. We use AI in our work because it helps builders and planners run better projects, faster, which is exactly what Sitedrive exists to do.
Our philosophy: AI doesn’t replace thinking, it accelerates it.
That quote came up in our discussion, and the team collectively nodded. AI lets us:
- turn thoughts into prototypes fast
- test ideas while they’re still fresh
- iterate before we overthink
- uncover flaws earlier
- understand systems more deeply
- remove friction from collaboration
The thinking still must happen. AI just moves it from the abstract to the concrete at record speed.
What’s next
We’re not chasing the next flashy AI trend. We’re refining how we build, improving our workflows, sharpening our engineering practices, and learning what produces the biggest customer impact.
“There’s a lot of untapped potential still. The tech is already strong, now it’s about our own processes,” Jussi says.
And as we keep building the future of construction production management, AI is becoming not just a tool, but a multiplier of what our people already do best: solving real problems for builders.
