<- Back to home
ToolingAI

Zürich AI | AI Tooling

February 24, 2026 17:30 ETH AI Center, OAT X11, Andreasstrasse 5, 8092 Zürich Panos Vagenas + Patrick Jayet + Rob Smith

Hands-on AI insights, practical learnings, and time to connect with the community.

"Learn how to fully utilize unstructured data in AI workflows."

Speakers

Panos Vagenas headshot
Panos Vagenas
Docling Technical Steering Committee Member
IBM
Patrick Jayet headshot
Patrick Jayet
Global Data & AI (GDAI) Innovation Unit
AXA
Rob Smith headshot
Rob Smith
Application Lead
Apertus at SwissAI

Details

We’re excited to welcome Panos Vagenas as our first speaker for 2026.

Talk abstract: Panos Vagenas

Join this session from the core Docling team to learn how to fully utilize your unstructured data in your AI workflows, be it for RAG, information extraction or, ultimately, any document-related agent. We will go over Docling’s main capabilities, covering how it can help address your enterprise needs, as well as giving a sneak peek of upcoming developments.

Talk abstract: Patrick Jayet

Earth Observation for Enhanced Risk Awareness of Commercial Properties

AXA insures millions of commercial properties worldwide. To strengthen risk awareness across its insured portfolio and streamline Risk Engineering workflows, AXA’s Global Data & AI (GDAI) innovation unit has created some time ago a new initiative combining Earth Observation and computer vision. This talk will describe how the project was initiated and developed, and how it supports risk engineers by providing scalable, data-driven insights for commercial property risk assessment.

Talk abstract: Rob Smith

From Models to Systems: Running and Applying LLMs in Practice

Apertus is a fully open-source, multilingual foundation model built for open and reproducible AI. This talk introduces the model and shares practical lessons from deploying Apertus and other state-of-the-art models for internal users. The focus is on turning model weights into usable systems.