Oct 30, 2025 — article

Takeoff Tuesday with Cognite’s Petteri Vainikka: Bridging the Gap in Industrial Data

IMG 0958 1

By Mari

At this month’s Takeoff Tuesday, we welcomed Petteri Vainikka, EMEA President at Cognite, for a sharp and candid look at where industrial data and AI are heading and why the winners will be the ones who experiment faster, not perfect longer. 

The state of industrial data: promising, but still lagging

While consumer tech has raced ahead with AI at every turn, industry continues to wrestle with legacy systems, siloed data, and slow adoption. As Petteri put it, “Most factories aren’t even fully connected yet, but that shouldn’t stop progress.” 

Instead of waiting for perfect data harmonization, Cognite pushes companies to contextualize what they already have, linking IT, OT, and engineering data into a knowledge graph that AI can actually understand and act on. 

Bridging the gap between hype and reality

IMG 0942 1

Petteri’s “State of the Union” wasn’t a future-gazing presentation. It was a grounded look at how Cognite has evolved over three years of market testing, iteration, and honest failure. 

He described the gap between industrial reality and AI ambition as a “Grand Canyon,” bridged only through relentless testing and transparency. His message: 

“Jewelry sells better than shovels. People don’t want tools, they want value.” 
“Agents are the new APIs. They’ll let systems talk without endless integration.” 
“Speed is the only moat. The only defensible edge is execution.” 
“Everything is a best guess. Test, fail, learn, and go again.” 

It’s a mindset that fits right into the Runway ethos: build fast, learn faster, and bridge bold innovation with operational reality. 

On AI skepticism and trust

IMG 0946 1

An audience question captured the elephant in the room: “What do we do when people in our business don’t trust AI results when accuracy isn’t guaranteed?” 

Petteri’s response was simple and honest. AI will never be perfect; it’s a probabilistic “dream machine,” not deterministic math. The key is to design for verification, not blind trust. Show users where results come from, give transparency into the data sources, and make every answer traceable. 

In other words, don’t wait until it’s flawless. Engineer trust step by step. 

Petteri’s final lesson echoed through the room: throw things at the wall and see what sticks. 

Don’t wait for certainty. Don’t overanalyze. Move fast, test ideas, and learn from what fails, because execution speed is defensible. 

 

 

IMG 0930

At Runway, we’re building the platform where founders scale and industry innovates. Through our VC fund, Tech Hub, and Industry Software Alliance, we back the next generation of industrial software startups using data and technology to make industry more efficient, sustainable, and reliable. 

IMG 3103 1

Every Takeoff Tuesday ends in our tech hub with great sandwiches, strong coffee, and deeper conversations about what’s next. We hope you’ll join us for the next one 💌

Be the first to receive our latest news