How Swiss Startups are Making Their Mark on Agentic AI

Agentic AI may be new, but the market is already crowded and those who want to succeed must have something special to offer. Swiss AI startups have such specialist services thanks to their entrepreneurial experience and scientific background.

Agentic AI refers to AI systems that are able to pursue goals without requiring human intervention. The hope is that these digital co-workers will be able to perform routine tasks independently. Although the technology is young, the market is already competitive. Those who want to compete for customers and investors require AI agents that are particularly innovative, easily integrated into existing systems, or particularly well adapted to specific task areas.

This is clearly the case with Zurich-based startup Unique. It is at the forefront of agentic AI innovation for asset management, wealth management and retail and private banking. Unique AI is deployed by blue-chip companies with more than USD 2.3 trillion in assets under management, including Pictet Group, UBP, LGT Private Banking, SIX and other top players in the financial services space. One of Unique’s most prominent clients, Pictet Group, offers the platform to 6,000 employees and reports efficiency gains of about two hours per week per person. With this track record, investor interest is keen. At the end of February, Unique closed a series A round of USD 30 million.

Unique is backed by an experienced team: founders Manuel Grenacher and Andreas Hauri are serial entrepreneurs and sold their last company to SAP. Swiss agentic AI startups are also able to draw on experience from big tech; for example, the two founders of Hyperion AI worked for many years in Google’s Zurich office. Hyperion AI was recently selected for the IMD Startup Challenge, one of the competitions on our Channels to Watch list.

In Sonar’s case, the experience differs: the Geneva-based unicorn, a leading provider of code quality and code security solutions, acquired AutoCodeRover, an autonomous agentic AI platform for software development, in February.

At LogicStar, entrepreneurial experience meets academia. The startup, which raised USD 3 million in a pre-seed round, has built an agentic AI for fully autonomous application maintenance to make self-healing software a reality. Three of LogicStar’s co-founders previously built DeepCode.ai, a technology trusted by millions of developers (acquired by Snyk and now called Snyk Code), and scaled it to more than USD 100 million annual recurring revenue (ARR). Co-founder Martin Vechev is a professor at ETH Zurich and heads the university’s Secure, Reliable, and Intelligent Systems (SRI) lab. According to the current Times Higher Education World University Rankings, ETH is one of the five best universities worldwide in computer science, so it is not surprising that other spin-offs in the field of agentic AI have emerged.

In mid-March, ETH spin-off DeepJudge launched a suite of AI agents and applications for law firms allowing instant, accurate access to all a firm’s internal data, and thus enabling customers to build AI-powered workflows tailored to their specific needs. The product, AI Workflows, was developed in close collaboration with customers: Lenz & Staehelin, one of Switzerland’s leading law firms, played a pivotal role in shaping it and other customers are now using AI Workflows.

The founders of nunu.ai also met at ETH Zurich. They develop AI agents for an application outside the mainstream, so they face less competition. Their digital co-workers test video games and since the launch several gaming studios, both large and small, have used the AI ​​agents to QA test their games, running hundreds of automated tests monthly. In March, nunu.ai secured USD 6 million in a seed funding round; investors included a16z Speedrun.

Find out more on why Switzerland is producing so many promising AI startups on our Innovative AI focus sector page.