Blog · Nov 19, 2026 · 8 min read
Slush 2026: Don't Fear the Agents. Hire Them.

A dark, wet and very bright week in Helsinki
Hannu Verkasalo, CEO & Co-Founder, Behavix · Helsinki, 19 November 2026
I am back in Helsinki for the first time since the summer, and the city has done what it always does in November: switched off the sun, turned the rain sideways and invited the whole tech world over anyway. Slush has never apologized for the weather. It built a brand on it! 🙂
My first Slush was in the very early days, I think around 2010, probably the second or third edition, when it was some 250 people and you could meet all of them before lunch. This year it is 12,000+ people (and like rumored in the city, likely around 10,000+ extra people not attending the event, but anyways in Helsinki to network…), about 5,800 startups and 3,300 investors, plus more than 600 side events around the city. And the funny part… the organizers cut the audience by 10 percent on purpose, to serve early-stage founders better. Only in Finland do you grow by shrinking. 🙂 It is still a not-for-profit run by students and volunteers, the Slush 100 winner now takes home €500,000 with no equity taken, and the theme of the year is a bet on Europe. The same week ICEYE, born at Aalto University, made a €2.05 billion offer for the rocket maker Avio. Finnish startups now go shopping for launch companies. That tells you how far this ecosystem has come.
What I heard on stage
President Alexander Stubb opened the main stage in a fireside chat with Bloomberg’s Oliver Crook. Geopolitics took its share of the time, as it must. What stayed with me was how strongly he feels about Finland as an AI-native place to build, with real successes already, and his view that we are entering a transition of the same order as the industrial revolution. And I very much liked his statement, when asked about how Finland lost Nokia, missed the growth of smartphones and Internet’s consumer-facing services era, missed the first generation of AI revolution, he bluntly argued that this is all nostalgia… we need to look forward, focus on new opportunities, things that you can influence, do not look back… and I agree. And back to his original comments about AI, I would add one thing: the winners of the industrial revolution were not the people who owned a steam engine. Everybody had one soon enough. The winners had something worth powering.
Then Reid Hoffman, founder of LinkedIn and one of the earliest backers of OpenAI, sat down with Marko Ahtisaari for a talk on strategic optimism about AI. Reid’s eternal optimism and fast thinking inspired me, once again (seen him many times on stage). The best part came at the end: we should embrace AI as our agents, our agencies. On all levels, also when we study, and when we sit on boards.
I have used the same metaphor for years. Back in the day, if you were lucky, you had brilliant colleagues, assistants and experts working with you. You gave them the context, the goals, the strategy and the timeline, and they delivered, properly and cleverly, usually better than you would have. (That last part takes a founder a few years to admit.) Brands did the same when they handed their campaigns to media and ad agencies. Now the agency is software. It does not take over. It works for us, and our job is to brief it well.
Erik Bernhardsson of Modal gave me the line of the week: “If you want to start a billion dollar company today, build a data company!” Sounds familiar here at Behavix… 🙂 Software and products are already commoditized. Our Chief Architect Mika Ristimäki always says in our internal meetings that long-term, defendable uniqueness comes from only three things: hardware you control, unique proprietary data, or your own users and distribution. We have spent the last two years building the second one, while embracing great partners who master the third one…
And Benedict Evans, whom I much admire and follow, was excellent with his “AI Questions”. Two simple points stuck. First, AI lets us do things cheaper and quicker, which is true and has its own implications. Second, the level of abstraction of our questions goes up. I see this every week. In our industry the key questions used to be: who is using my app, is it growing or declining, what kind of people are they? Now the questions are: what is my churn, what drives it, and what should I do about it given what my competitors are doing?
Meanwhile, at the Behavix Center of Excellence
Before Slush, the week started with an on-site with our quickly expanding Helsinki team, the Center of Excellence for our Behavioral (Data) OS. Like always, we played around with cool new ideas, went through exciting new customer projects and integrations, had a good Christmas lunch (in Finland the Christmas party season starts when the darkness does) and took a good step ahead in planning next year. Great meeting the whole team again! 🙂
The rest of the week was 30+ great investors, many old good friends, angels, serial entrepreneurs and new founders. Dinners, private parties, cocktail events, morning coffees and brainstorms. My calendar looked like a game of Tetris played by somebody who was losing. Lots of activity, great vibes and spirits!
Three companies worth watching
Slush week is also when I catch up with the companies in our own corner, market research and insights. Three of them deserve kudos, each changing the industry in its own way.
Valogent. It was nice to meet the whole founding team after our initial talks, Jesse Sandqvist, Tomi Laitila and Markus Kuusisto. They do voice-AI moderated qualitative research at scale: an AI interviewer runs the conversation, the analysis is there right away, and the cost of a good interview drops to a fraction. They want qual to move from one-off projects to always-on infrastructure. I like where this meets us: behavioral data tells you what a person did, and their interviewer can ask the same person why, the same evening.
Viomba. Markku Mäntymaa and his team have measured human attention for about a decade, and I know Markku even longer (before embracing start-ups and media, he managed the legendary band Leningrad Cowboys! 🙂). Their models are trained on eye-tracking data from real people in real media environments, and the message is simple: an ad can be viewable and still unseen. They now turn that into tools that score and improve creatives and campaigns before the money is spent.
MyTelescope. Nice to have coffee with Rodrigo Graviz again! Their agentic market-intelligence platform reads demand from what millions of people search for across thirteen digital platforms. We announced our collaboration this autumn to bring behavioral data into the same layer, and the clients are already asking for more countries. A good problem to have.
And these are only three. There are many more interesting start-ups around here, and the Nordic ResTech scene is very much alive.
Where Behavix fits
Everything I heard this week lines up with four things I believe about our industry.
- Behavioral data is the core. We observe rather than ask. Opt-in, permission-based, on people’s own phones and computers. What people actually do is the ground everything else stands on.
- Synthetic data is not enough. It is useful for stretching a real sample and for exploring ideas. It cannot invent behavior that nobody observed. Simulate when you explore, measure when being wrong is expensive. Multiply by zero and you still have zero.
- Independent measurement and competitive intelligence are the key. Every company has a perfect dashboard of its own product and a foggy window into everybody else’s. I talked about this with people with a Supercell background: a games company knows every tap in its own game and very little about what the same player does in the next one. The same goes for what people are shown in their feeds and apps before they decide anything. You only see that from the user’s side of the screen.
- Agentic AI is how it all gets used. Through our MCP, a client’s own AI assistant can query our data directly: competitive tracking, insights, consumer journey models, media and audience research, close to real time, without delays and without hefty investments. That is Benedict’s cheaper and quicker. And it is Reid’s agency: we do not only give access to unique data, we provide the tooling and agents to act on it.
So… the President says a revolution is coming. Erik says build a data company. Benedict says the questions get smarter. Reid says hire the agents. Put together, that is roughly our business plan. We did not coordinate on this when Behavix was put together in 2022… I promise. 🙂 Nobody conquers the world in a week, not even in Slush week. But we are well on our way in our corner of it.
Finally, thank you to all the great investors who hosted incredible events this week. I enjoyed it a lot! Special thanks to our investor Vendep Capital. See you next November. I will bring an umbrella (again)…

Co-Founder & CEO of Behavix