Dear readers,

This week’s edition plays out like a systems test for resilience: starting in airport transit zones and temporary shelters, and ending somewhere much closer to the future of AI.

At the center is a Romanian founder whose path doesn’t follow the usual startup narrative. It moves through instability, creative detours, and into some of the more complex questions in tech today: memory, autonomy, and what makes machines feel consistent over time.

Across Southeast Europe, capital is moving and new deals are taking place. However, as our investor of the week puts it, the harder question is shifting upstream: not how to build, but what’s actually worth building in the first place.

It’s a week about misfits, memory, and the systems - both human and artificial - that hold everything together.

Enjoy the read!
Bojan Stojkovski
Editor-in-Chief, IT Logs

For a while, Timmy Ghiurau lived in an airport. Not as a metaphor, or a founder’s origin story written for conference stages, but literally. Timmy was sleeping on airport benches during the day, and at night he was playing music in blues bars across Copenhagen just to get through another day. 

“I was homeless a few times in my life. Sleeping in the airport for three months, playing music wherever I could.” At that point, there was no sense that this would turn into anything resembling a career. He had left Romania to study music… at least, that was the version he believed in. 

From music to machines

The move into tech didn’t come as a clean pivot either. Here, it was Timmy’s curiosity that played a big role. “I was always interested in how you can augment creativity with technology. How you can build worlds, simulate things.” he tells IT Logs. 

That curiosity pulled him into real-time 3D and early work with Unity, at a time when the company was still expanding beyond gaming into simulation and research. “I joined when they were exploring verticals outside games,” he says. But what kept him there was an underlying question. 

Eye tracking doesn’t just improve interfaces - it reveals attention, intention, and cognition. “You see where people look, how long they look, what they ignore. You start to understand attention in a very granular way.” he says. 

It was an early encounter with a theme that would follow him: technology that’s moving a lot faster than the systems designed to govern it.

The Volvo decade

His time at Volvo Cars brought structure, but also a different level of complexity. He spent nearly a decade there, working across XR, AI, and innovation strategy, operating between ecosystems like Tel Aviv, San Francisco, and Stockholm. “You have to make things work in the real world,” he says. 

Over time, his role shifted from building prototypes to modeling futures. “I moved into future scenarios. Looking at signals, trying to understand what’s coming next.” he recalls.

The questions became broader, more systemic. “What happens when autonomy actually works? What does that change in behavior, in society?” It was no longer just about vehicles. “What happens when a car becomes an experience?” he says. “Or part of a lifestyle ecosystem?” That shift in mindset, from building features to anticipating systems, changed how he approached everything. 

“You start thinking on a macro level. You are thinking about governance, policy, or infrastructure.”  he tells IT Logs.

MidBrain, the startup he co-founded, began purely as a conversation. There wasn’t a business idea behind it. “I was at an MIT hackathon. Super jet-lagged.” he tells IT Logs. There, holding a high-end VR headset, he met his future co-founder, and they started debating a deceptively simple question: what makes a human real? “It’s not just photorealism, it’s having coherent behavior over time.” Timmy says.

Solving the problem of memory

That idea became the foundation of what they would build. “We are like a continual learning lab. Working with memory and learning infrastructure for agents.” he says. They started experimenting in Minecraft, using it as a sandbox for building autonomous agents that could operate over long periods.

Until one moment reframed everything. “I asked one agent, ‘Let’s go explore a cave,’” he says. “And it said no, it wanted to keep building a house.” He pauses, almost amused. “That’s what your friends do,” he says. “They don’t just obey you.” But that creates tension. “If I’m paying for an AI, I expect control. But if it behaves like a human, it needs autonomy.” he says. 

The company has since expanded into robotics and broader AI systems, but the focus remains the same: building agents that can learn continuously, adapt over time, and form something closer to trust. “Ultimately, you want agents that anticipate your needs and that understand your context.”

Where execution exists without ideation

Despite working across global ecosystems, Timmy keeps returning to Southeast Europe. “I’ve been more present in East Europe,” he says. “Traveling, talking to founders, funds, accelerators.” What he found isn’t a lack of ambition. “There is capital, for sure, more and more funds are emerging.” he says.

But something is missing, Timmy explains. “The first part of the double diamond is missing.” 

There are structural gaps too, such as weak connections with universities, investor frameworks that don’t fit deep tech, and a shortage of experienced angel investors. “What we’re missing is a critical mass of angels. People who can take risks early.” he tells IT Logs.

Having spent time in Sweden and Finland, Timmy often points to a different dynamic. “It’s like the PayPal effect. You need founders who succeed, then reinvest, then build again.” he points out. That loop creates momentum. “In Sweden, Finland. you see that,” he says. “Capital, experience, networks, all of these compound.” 

At the same time, Timmy does see a broader change happening globally. “Software is getting commoditized,” he says. “With large language models, anyone can build something.” That changes the rules. “You don’t need to code as much, but you need to understand what to build.” Which means differentiation moves elsewhere. “You need good design, good user experience - you need taste.” Timmy remarks.

The misfits and the long view

All of this creates an opening for Southeast Europe. “That’s why people are more receptive now,” he explains. “To learn how it’s done in the Nordics or elsewhere.” There are also macro shifts, talent moving out of traditional industries, there’s a new interest in dual-use technologies, increasing cross-border collaboration. “There’s a new flavor here, and investors are starting to notice.”

The problems he’s working on, such as AI, memory, human behavior, don’t belong to a single discipline. “I like to go in areas where people are not usually exploring, and try to push it.” Timmy points out.

When he evaluates startups, or his own work, he reduces it to a simple question. “How does the world look with this, and how does it look without it?” It’s a way to strip things down to what matters. 

That question applies just as much to SEE as it does to AI. What happens if the region keeps optimizing for execution? And what happens if it shifts toward ideation, design, and long-term thinking? “That’s the gap, and that’s where the opportunity is.” 

The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.

Learn more

Got tech rumors for us? Reach us at [email protected]

IT Logs: Where do you think the smartest money is going this year?

IT Logs:  Bigger bets on fewer startups, or smaller checks spread wider?

IT Logs:  What is moving valuations more in 2026: real traction, AI buzz, or solid unit economics?