What early experiences shaped your entrepreneurial journey?
I grew up in a family deeply rooted in innovation—my grandparents ran a precision-mechanical company that became IBM’s first gold supplier in Europe. They played a role in that era of innovation, manufacturing fine mechanical elements integral to the mainframes that were early computers. My grandfather founded the company before World War II, and it was a testament to his resilience and entrepreneurial spirit. His legacy shaped my own journey as an entrepreneur.
Your initial studies were in economics before you did a PhD in Information Systems. Why this transition?
I did a degree in econometrics at the University of Geneva, and I was involved with a student organisation which gave me the opportunity to study for a PhD. So, I went to Aberdeen in Scotland and then completed the PhD back in Switzerland. I realised that spending my days producing economic research papers to be read by no one but myself wasn’t for me.
How did you start your career?
While I was studying for my PhD we were literally wiring the first web pages for companies like UBS. From this, some partners and I founded a spin-off company called Namics which grew into the largest business transformation agency in Switzerland and later expanded into Germany. We started with three founders and scaled the company to employ 800 people, designing websites often for high profile clients.
Towards the latter stages of Namics, myself and one of my clients co-founded our next company, localsearch.ch, a search platform, which digitised the entire yellow pages of Switzerland. This company had a straightforward business model based on small transactions which, when scaled, amounted to a huge business. After taking it to market leadership within four years, even surpassing Google for local search, we sold localsearch to the telecoms giant, Swisscom.
As a serial entrepreneur, have you suffered any setbacks?
Yes, after local.ch, I co-founded Memonic, a note-taking application but unfortunately, our value proposition was almost exactly the same as Evernote’s. Despite the strength of the product and its popularity, we found raising venture capital funding very difficult. It was a freemium product which struggled to get paying users and this just wasn’t sustainable in the long-term. We learnt a good lesson and out of the ashes, Squirro was born. This was based on what we had done previously, adapted to enterprise search.
Why did the market at that time need a service like that offered by Squirro?
I looked around and realised that enterprises were drowning in unstructured data (emails, documents, call notes, etc) but they were lacking context-aware insight. I saw that there was an opportunity to use AI-powered technology to filter and surface relevant information precisely when it was needed. No company wants everything all at once. Amidst all our experience in search and consultative services I realised we had the ability to make sense out of this enterprise data chaos.
How does Squirro leverage AI, and what sets its technology apart?
Squirro’s core offering—the GenAI platform—takes organisational data from any source and using AI, turns it into actionable insights that can help to drive informed decisions. At the centre of this is an insight engine which merges machine learning, semantic search, and generative AI to analyse structured and unstructured data streams, surfacing contextual knowledge and the “why” behind the data. It also integrates knowledge graphs and taxonomies, enabling Retrieval-Augmented Generation (RAG) with factual accuracy and traceability.
I can’t emphasise enough that generative AI must be reliable—Squirro enhances LLM outputs by grounding them in enterprise-controlled data, minimising hallucinations and boosting compliance.
Why is precision and trust so critical for Squirro’s target sectors?
We focus on highly regulated industries—banking, insurance, logistics—where compliance, accuracy, and traceable AI are non-negotiable. Our team builds with regulatory standards in mind: strict validation, evidence-based results, and enterprise-grade security to earn trust from exacting institutions like the European Central Bank, Standard Chartered, ING and HM Revenue and Customs.
What key pitfalls should enterprises avoid when launching AI initiatives?
There are several considerations that all organisations should be aware of. The first is siloed data landscapes. Most enterprises have hundreds of apps, only 30% of which are integrated. This results in unstructured data amounting to 90% of all enterprise data, making it easy to miss critical insight. Even with all the publicity surrounding LLMs (Large Language Models), these are still hard for businesses to understand and use appropriately. While generative models excel at creation, they can hallucinate, so enterprises need to combine them with structured knowledge and retrieval mechanisms. It’s also important to remember to keep humans in the loop. AI should augment, not replace, human decision-making, and I like to think of AI as a partner rather than a replacement. Finally, when it comes to governance and regulation, addressing security, privacy, and compliance upfront avoids downstream risks which are especially important in regulated industries.
What would you say are the skills needed to be an entrepreneurial success in AI?
Curiosity and boldness with an “insane” obsession for solving real problems, not just tech-for-tech’s-sake! You also need resilience and reinvention, like I had after the failure of Memonic, the Evernote-like app, where you must be able to pivot to new opportunities and learn from past experiences. It helps to have domain expertise, understanding both technology and industry nuances enabling tailored, enterprise-grade solutions; and it’s helpful to have a value-first mindset, which means focusing on delivering measurable ROI for your customers.
