Into the wild - AI in Business
Circle - Convention Center
Airport Zurich
17 November 2026
Keynotes GTC26
Leading Independent Tech Analyst
Benedict Evans
What matters in tech? What’s going on, what might it mean, and what will happen next?
Distinguished Engineer at Google Cloud
Alon Halevy
Data Management Meets its AI Partners
Leaders Track 1
10:30
Benjamin Theunissen
Head Data & AI
Helvetia
-
Stories of use cases, scaled applications and business transformation. A merger full of opportunities and insights. AI is reshaping insurance by improving customer service, increasing productivity, and elevating the importance of human judgment, curiosity, and critical thinking. Helvetia Baloise’s AI strategy highlights the shift from task-based optimization to agentic, end-to-end process automation, supported by reusable group capabilities, governance, and adoption. Practical examples, including AI-supported claims handling and agentic software delivery, demonstrate how proven AI foundations can scale into measurable enterprise value and long-term transformation.
Biography:
.
11:10
Gaudenz Koeppel
CDAO Trading & Sales
Axpo
-
Abstract will follow soon
Biography
Biography will follow soon
Lunch 11:50 - 13:00
.
Leaders Track 2
10:30
Attila Steinegger
Head Digital Transformation & Data
WWF Schweiz
-
Climate change, biodiversity loss and the transition towards a more sustainable economy require greater transparency and faster decision-making. At WWF, data and artificial intelligence have become strategic assets for understanding these systems and identifying opportunities for change.
In this session, Attila Steinegger shares how WWF leverages data and artificial intelligence to support conservation and sustainability goals, both internally and across markets, value chains and regulatory ecosystems. Using the example of Assessfy, WWF's AI-powered assessment platform, he will demonstrate how AI can transform large volumes of complex sustainability information into actionable insights that help organisations identify risks, opportunities and improvement areas.
Beyond the technology, the session offers practical lessons on building impactful AI solutions in a mission-driven organisation and explores how transparency can become a powerful lever for driving system change at scale.
Biography
Attila leads the digital transformation of WWF Switzerland and drives the strategic adoption of data, artificial intelligence and emerging technologies across the organisation. His work focuses on turning innovation into measurable impact, enabling WWF to make better decisions, work more effectively and accelerate its mission. Previously, he spent a decade in consulting, focusing on strategy, innovation and digital transformation. In 2023, he was named as one of the Top 100 Digital Shapers in Switzerland by Bilanz. He holds a bachelors degree in Business Administration from the University of Applied Sciences St. Gallen and a masters degree in Corporate Sustainability from Cranfield University in the UK.
11:10
Ueli Zumkehr
Learning Experience Design
Die Mobiliar
Sebastian Baumhauer
Principal Consultant
D ONE
-
Abstract will follow soon
Biography
Biography will follow soon
Lunch 11:50 - 13:00
Into the wild - AI in Business
Once again this year, the GTC is offering two Leader Tracks where leading figures will share their insights. Authentic and valuable.
13:00
Anne Thielen
Director Data & AI Solutions
Sonova
-
We are transforming hearing aid reliability by using artificial intelligence. Through innovative system screening we determine reliability indicators in real-time.
Biography:
Since 2025, Anne holds the role of Director Data & AI at Sonova’s R&D. She earned her Ph.D. in Applied Solid State Physics from ETH Zurich. Following this, she focused on transforming hearing instruments into body-worn IoT devices within Sonova’s R&D department. Her current interests include real-world clinical grade outcomes, classifier integration, and servitization.
.
13:40
Johannes Dueck
Head of AI Assets
Ergo
-
Abstract will follow soon
Biography
Biography will follow soon
Break 15’
14:35
Mark Schwarz
Chief Digital & Information Officer
mobilezone AG
-
How do we move from AI pilots to lasting business value? In this talk, I will share mobilezone's approach to connecting data, AI and business priorities, and discuss where agents can support decisions and day-to-day work. I will cover the choices behind prioritisation, ownership and governance, as well as the practical questions teams face when scaling use cases. Rather than focusing on products, the session will offer lessons on what it takes to turn promising ideas into useful, responsible outcomes—and where we still need to learn.
Biography
Mark Schwarz is Chief Digital & Information Officer at mobilezone AG. He leads Digital & IT, bringing together digital engineering, data, AI and IT capabilities. His work focuses on how technology can support business priorities and practical decision-making. At the Growing Together Conference, he will share his perspective on moving from AI ideas to useful applications in day-to-day operations.
15:15
Eric Gravil
Chief Data & AI Officer
Vaudoise Assurances
-
Few of us get to start on a blank page. Most inherit a platform, a decade of undocumented logic, and a business that depends on both. Getting funding is hard, but it's the part people prepare for — what decides the outcome is the string of decisions between the go-ahead and the day you switch the old system off.
BiographyEric is Chief Data and AI Officer at Vaudoise Assurances, where he leads the data organization — from governance through the data platform to AI — and is responsible for defining and delivering the company's data strategy. Twenty years in regulated industries, pharma and insurance, have taken him from strategy and vision to hands-on delivery.
.
13:40
Jonas Stehli
Head Quantitative Analytics Pricing & Revenue Development
Bank Julius Bär
-
Abstract will follow soon
Biography
Biography will follow soon
.
13:00
Tamara Hofer
Leiterin Data, Analytics & KI
Luzerner Kantonalbank
-
Abstract will follow soon
Biography
Biography will follow soon
Break 15’
14:35
Marco Fedeli
Head Data & Digital Experience
Emmi
-
Transforming milk into cheese requires more than technology. It demands craftsmanship, patience, trust, and a clear vision of the value to be created. The same is true for data transformation. In this session, Emmi shares its journey to a modern data platform and the challenges encountered along the way: overcoming resistance to change, establishing ownership, balancing governance with self-service, and building a data-driven culture. An honest leadership story about why the biggest challenges are rarely technical, but human.
Biography
Marco is Head of Data and Digital Experience at Emmi, where he leads the company’s data, analytics and digital workplace capabilities. With a background in Business & Economics and over 10 years of experience in IT, data and business transformation, he combines strategic leadership with a strong hands-on mindset. His passion lies at the intersection of business, technology and data – turning complex challenges into practical solutions and using data to enable better decisions.
.
11:10
Data & Platform Engineering
Bernhard Vennemann, D ONE
Fabian Forrer GKB
-
Bringing a modern data and AI platform to life in a regulated environment demands tight alignment between rapid innovation and strict enterprise control. Teams need to deliver tangible business value quickly while ensuring enterprise-grade security and compliance from day one. As new AI workloads stretch the limits of traditional infrastructure, organizations need a pragmatic approach that avoids both lengthy, over-engineered delivery cycles and fragile, shortcut-driven builds.
13:55
Analytics & Insights
Alexander Stanenko, D ONE
Luqman Chettiyarthodi, ERGO
-
Every production RAG system presents the same difficulty: dozens of parameters across retrieval, generation and prompt, interacting in ways that manual review cannot resolve. The result is that systems get improved without anyone being able to say how good they are, or how much better they could be.
KCSP is ERGO's assistant for 6,500 sales partners, answering product questions whose content reaches customers. This talk follows its path from that position to a state where configuration is a measured decision rather than a matter of judgement, and what it took to get there: a definition of quality the business recognises as its own, and the means to evaluate a configuration against it without anyone reading the answers.
The outcome was a measurably better assistant, and a way of arriving at the right configuration before deployment rather than after.
10:30
Data & AI Governance
Heiko Kromer, D ONE
Andreas Ambühl, CSS
-
At CSS Health Insurance, claims processing is the heartbeat of the business: high volume, every day, fast and reliable. We wanted data access management to work the same way. Instead, it ran at the speed of JIRA tickets. Because granting access required writing Terraform and using GitOps processes, data owners had to wait a long time for their data access to change.
This talk shows how we built a scalable, automated metadata-driven access management service for Azure Databricks that treats access requests like claims: declared in a human-readable YAML file, processed automatically, and auditable end to end. GitHub Actions and Terraform perform the work behind the scenes and hidden from the requester. The service covers the Databricks workspace, Unity Catalog and Power BI. It is IAM-driven and gives data owners real autonomy within clear guardrails in a highly regulated industry.
We'll cover the business case, architecture, tech stack and end-to-end process, then give an honest review of what worked and what did not. We close with an outlook: when all your access metadata lives in YAML, it's exactly what LLMs handle well at volume.Description text goes here
11:10
Data & Platform Engineering
Marco Petrella, D ONE
-
abstract will follow soon
10:30
Analytics & Insights
Fabian Müller, D ONE
Ivan Smurov, ERGO
-
abstract will follow soon
13:55
Data & Platform Engineering
Andrei Dmitrenko, D ONE & Gabriel Levaillant, D ONE
-
A year ago, running a capable language model on your own hardware and getting it to do meaningful work was out of reach. Today we do it every day. An open-weight model on a laptop or an on-premises server can take on a real task, such as building and evaluating a machine learning classifier end-to-end, and finish it with no cloud, no API bill and no data leaving the machine.
In the first half of the talk, we show what this looks like in practice, drawing on a series of hands-on experiments with local agentic setups. We cover what open-weight, small language models can accomplish today, where they still break, and why the setup around the model, from the agent to the inference server, matters as much as the model itself.
In the second half, we turn to regulated environments. Drawing on our projects with Swiss hospitals in intent classification, retrieval-augmented generation and systematic evaluation, we discuss why locally deployed models change the equation for healthcare and other data-sensitive domains, and how we are building the capabilities to deliver them for our clients.
14:35
Smooth Operations
Alexandra Studer, D ONE
Andreas Spinner, Die Mobiliar
The Data & AI Tech Talks will follow soon:
As every year, experienced professionals will discuss applications and technical solutions from the top platforms. Here’s a sneak peek at some of the topics:
More topics to follow by the end of the month.
Framework for successful Data Initiatives
From promising idea to scalable data product.
____
Designing data platform
What a future-ready data architecture needs.
____
RAG Evaluation
How do you know if your RAG system is actually good?
AI E2E Automation Framework
From individual AI tasks to automated workflows.
____
Automated Data Governance
Can AI help govern data at scale?
___
Upskilling the AI organisation
Moving from individual experimentation to organisational capability.
___
Semantic Layer and Ontology
Giving AI the business context it needs.
___
Finops +AI
Keeping the cost of AI and cloud under control.
Small/Open Models
When smaller can actually be smarter.
___
LLM for information extraction
Turning unstructured documents into structured, usable data.