Turku: Finland’s hub of AI innovation

Share

The sleepy medieval town of Turku in Finland is becoming a hub of deep tech, where thorough academic research meets the latest in AI innovation. Chris Price reports 

It’s probably fair to say that, outside of Finland, very few people have even heard of Turku. ‘You mean Turkey,’ one of my friends said when I announced I was heading there on a five-day press trip. Yet Finland’s oldest city and former capital – until the Great Fire of Turku destroyed most of its buildings in 1827 – has huge ambitions to become northern Europe’s leading AI innovation hub.

As it prepares to celebrate its 800th anniversary in 2029, we were invited by Business Finland to visit the old city before heading two hours down the coast to Finland’s current capital, Helsinki for more yet more AI-related events: culminating in the two-day Slush tech event (more about that later).

Taking the plunge

First, at a welcome dinner taking place at Villa Marjaniemi (built in 1850 as a summer retreat for Finland’s Rettig family whose wealth lies in the tobacco industry), we were given presentations from several leading Turku tech entrepreneurs. Then, the evening ended as any serious Finnish business discussion should, with the heat of a traditional wood-fired sauna followed by a sudden, sharp plunge into the icy, dark waters of the Baltic Sea!

Remarkably, Finland has 3.3 million saunas for just 5.5 million people with most Finns, it seems, either having their own private saunas or sharing one with others in their apartment block. Statistics show that Finland has even more saunas than cars (currently there about 3.2 million vehicles on Finnish roads).

High-quality research 

Inevitably, the question for any European tech cluster is: just how does it compete with the scale and capital velocity of a US Silicon Valley or China’s Shenzhen? For this very small northern city (population approximately 200,000), the answer lies in leveraging its academic talent, particularly from the University of Turku.

At least that’s the view of one of Turku’s most famous sons, 39-year old Peter Sarlin. A Finnish AI entrepreneur with a PhD in Information Systems from TUCS (Turku’s Center for Computer Science), Sarlin founded his AI company Silo AI in Helsinki in 2017 before selling it to American IT giant AMD for approximately $665 million last year (the business is now called AMD Silo AI).

At the dinner, Sarlin, who is also a Professor of Working Life at Finland’s Aalto University and driving force behind the influential Turku AI Summit, highlighted the foundations of Finland’s strategic strength. “We have the right mix of computational muscle and theoretical grounding,” he noted, emphasizing that theoretical AI research is coupled with high-stakes deployment.

“Finland’s emphasis on high-quality research, combined with a willingness to tackle hard, industrial problems, makes us a critical node for developing safe and advanced AI solutions,” he added. “We’re not just building models; we’re building deep-learning architectures for the most demanding real-world applications, from logistics to health.”

This focus is a deliberate strategic choice, positioning Finland in high-barrier-to-entry sectors where precision, security and reliability (and not merely speed to market) are paramount.  For example, Sarlin’s firm is addressing complex, capital-intensive challenges that require long-term investment, moving beyond superficial AI trends.

“This deep tech focus ensures that the region’s AI output is globally relevant and structurally sound, moving beyond shallow consumer trends and into areas of true sovereign and industrial advantage,” he concluded.

CardioSignal: the battle for the pocket diagnostic

Of the several tech companies we met in Turku, perhaps the most compelling was CardioSignal. Founded by Juuso Blomster, it’s attempting to revolutionize cardiac diagnostics by converting the common smartphone into a clinically validated medical device capable of screening for life-threatening conditions.

The CardioSignal app leverages the phone’s existing high-precision motion sensors (accelerometer and gyroscope) to detect signs of atrial fibrillation (AFib) – a common, often hidden, and dangerous heart rhythm disorder responsible for strokes.

The methodology is simple and effective: the user places the phone on their chest while lying down, and the app analyzes heart movement data (ballistocardiography) with clinically proven accuracy in under sixty seconds. This simple act aims to transform heart monitoring from an occasional inconvenience into a frequent screening tool.

Blomster was certainly bullish about CardioSignal’s ambition. AFib is merely the point of market entry. The company has a significant pipeline aimed at tackling other critical, large-scale heart conditions.

“Heart failure is the only heart disease at the moment that is still increasing,” he said. “All other heart diseases are decreasing, but heart failure is increasing, and we’re addressing that,” he explained, identifying a major crisis zone in Western healthcare systems.

The pipeline extends to aortic stenosis, the most common valve disease where late diagnosis costs lives. Blomster addressed the systemic failures directly: “Even in the United States, with the most expensive healthcare system in the world, half of patients with severe aortic stenosis die before they get treated because the diagnosis comes too late,” he said. “We can change that.”

Regulatory friction?

Blomster’s strategic vision is to create a digital platform economy model for certified healthcare screening. “Let’s do what Uber does in transportation. Let’s do what Wolt does for delivery services. Let’s do that for the healthcare industry,” he declared, pushing a model of disruptive efficiency into a highly regulated sector.

However, the path is fraught with significant business and regulatory friction. The first challenge is overwhelming competition. Global tech giants like Apple and Samsung are aggressively pushing their own health monitoring capabilities, leveraging massive installation bases and deep pockets to integrate diagnostics into smartwatches and phones.

To succeed, CardioSignal must compete directly with ecosystem lock-in and consumer trust which is already established by these giants, making rapid, widespread consumer adoption difficult without powerful partners.

However, perhaps the more significant hurdle is regulatory approval and clinical acceptance. As we highlighted to Blomster, there is a profound difference between using your phone to book a taxi and using it to spot a heart condition.

Blomster confirmed that this risk is mitigated by adhering to strict medical device standards and bypassing the volatile direct-to-consumer market. The technology is not being sold as a simple consumer app but instead is being distributed via healthcare providers, clinics and hospitals.

“We have the connection so that it can be integrated into healthcare system services,” Blomster elaborated. According to him, ensuring CardioSignal is a B2B solution within existing clinical infrastructure is key to scaling adoption while navigating the immense regulatory barriers inherent in the medical space.

Deep tech: securing the sea and semiconductor

Of course, Turku’s AI capabilities extend way beyond medical technology. Given its location on the Baltic, a particular focus is on how tech can be used to improve shipping.  At the event, Karno Tenovuo, CEO and Co-Founder of Awake.AI, detailed the company’s mission to impose predictive intelligence on the global maritime sector – an industry notoriously resistant to digital change and ripe for optimization.

The Finnish company provides an AI-driven platform for smart ports and shipping, focusing on maximizing efficiency. Its immediate business case centres on optimizing berth planning, reducing vessel waiting times, and, crucially, cutting shipping-related emissions. The company’s target is a measurable 10% reduction in CO₂ emissions by 2030, linking commercial efficiency directly to global sustainability metrics.

From the University of Turku, Dean and Professor Petriina Paturi introduced the ambitious OCEANAut (On-Chip Edge AI Neuromorphic Applications for Security). Recently awarded 5 million Euros from Business Finland, the project is focused on energy-efficient, memristor-based neuromorphic computing.

This foundational research aims to develop a new, highly energy-efficient form of computing inspired by the biological brain. The goal is to move AI away from cloud dependency and solve the massive energy consumption issue.

“Current AI demands are unsustainable,” Professor Paturi stated bluntly. She noted that a single question posed to a modern AI takes ten times the energy of a basic web search. The goal, therefore, is to develop a new, highly efficient form of computing. This neuromorphic approach is inspired by the biological brain, which operates on just 20 watts.

By creating these brain-like chips, the team hopes to revolutionize the field. “By moving to new architectures like neuromorphic computing, we can reach the energy efficiency of the human brain – and that is the only way forward,” Paturi added.

Focus on defence

Finally, MarshallAI CEO Marcus Nordström, demonstrated its expertise in AI-driven signal processing for the defence, security, and manufacturing sectors where the automation of complex sensor data is vital.

Its flagship product, the Agile Focused Signal Analysis (AFSA) software, allows users to automate the analysis of complex images, audio, and radio signals. This can be done without requiring deep coding or AI expertise. This positions MarshallAI as a key enabler for legacy industries seeking to modernize and automate core security and quality control functions.

As Nordström concluded: “Our mission is to bring powerful, complex signal analysis out of the lab and directly into the hands of industry, today.”