Swiss Re hosts transformative event showcasing the future of open catastrophe modelling
Zurich, Switzerland – June 19, 2025
Leading professionals, insurers, reinsurers, and technology providers came together in Zurich for thehighly anticipated Oasis Insight Conference 2025. Hosted by Swiss Re, the event drew many attendeeseager to explore the latest innovations in open-source catastrophe modelling and collaborative riskassessment.
The conference opened with a warm welcome from Beat Aeberhardt, Head of Catastrophe and Geo
Modelling at Swiss Re, followed by opening remarks from Dickie Whitaker, Chief Executive at Oasis LMF.Both speakers emphasised the fundamental importance of community and collaboration within the fieldof natural catastrophe modelling, setting the tone for a day focused on shared innovation and opportunityfor advancing risk assessment capabilities.
Community-Driven Innovation Accelerates
Delivering the conference keynote talk, Dickie highlighted the Oasis LMF community now encompassesover 100 risk models from over 20 model providers. This isn't just growth - it's proof that opencollaboration can scale from academic concept to industry standard, supporting everything from localdeveloper installations to enterprise cloud deployments
The expansion of Open Data Standards (ODS) beyond property coverage represents a quiet revolution inprogress. While property standards have achieved widespread adoption, the active development ofliability coverage standards - with cyber and marine in the pipeline - signals something bigger thantechnical advancement. It's the creation of a common language for risk assessment.
The community's innovation projects are delivering tangible benefits: thorough model evaluationframeworks, global exposure model development, and academic collaborations. Whitaker concluded byhighlighting how technical working groups are actively addressing critical industry challenges, includingclimate change integration, multi-peril correlation frameworks, real-time data integration, and AI/ML bestpractices in catastrophe models.
Every building on Earth - Global Exposure Model
Senior Scientist at GFZ, Danijel Schorlemmer’s talk, presented the Global Dynamic Exposure Model. Acomprehensive initiative to create the world's most detailed database of buildings and their exposure tonatural hazards. Every building on Earth.
The platform is based on OpenStreetMap (OSM) ecosystem, using a crowd sourced approach for datacollection. The platform can integrate multiple open data sources with standardised taxonomies to
estimate detailed building-level information. Each structure in the database includes data on location,structural characteristics, occupancy patterns, and vulnerability assessments—creating anunprecedented level of granularity for global risk modelling.
Perhaps most significantly for the insurance industry, the platform supports multi-hazard exposure andloss calculations while being designed to serve both humanitarian applications and commercial usecases. This dual-purpose approach could improve accessibility to high-quality exposure data whilesupporting the industry's growing need for more precise risk assessment capabilities.
Flood Modelling Takes Center Stage
The following session on state-of-the-art flood modelling, featured experts Paul Della Marta (Head ofCatastrophe Research, PartnerRe), Petr Puncochar (Head of Flood Model Development, Aon ImpactForecasting), Malcolm Haylock (Head of Catastrophe Modelling, Fathom) and Jane Toothill (Head of CatModelling, JBA), who walked attendees through the current landscape of flood risk assessment. Thediscussion covered topics from data collection and hydrological modelling to the applications of machinelearning in flood prediction.
The panellists explored improvements in making uncertainties in the modelling visible, and challengesaround integrating climate change in an actionable way. The session highlighted the growing role of AIand ML in Impact Forecasting with examples ranging from traditional machine learning algorithms used incyclone track propagation to deep neural networks for high-resolution precipitation modelling. GenerativeAI applications are increasingly being used for coding support and documentation, marking a newchapter in modelling efficiency.
Professional Development Takes Priority
Peter Zimmerli (Senior Cyber Modelling Speciality, Aon) took to the stage to present the InternationalSociety of Catastrophe Managers (ISCM) approach to elevating the catastrophe modelling profession.With over 300 credentialed professionals now certified through their programs, ISCM is addressing theindustry's need for standardised competencies and recognised expertise.
The Credentialization pathway offers two distinct tracks: the Certified Specialist in Catastrophe Risk
(CSCR) for best practice experts, and the Certified Catastrophe Risk Management Professional (CCRMP)for recognised subject matter experts capable of advanced applications and critical assessments.
The organisations Coffee Talk series continues to bring key industry topics to experienced professionalsin accessible formats - mail campaigns will be delivered when registrations is available.
More information can be found on their website: ISCM | International Society of CatastropheManagers
Expert Panel Tackles Model Validation Challenges
An afternoon panel on "Model Validation and Evaluation" brought together industry leaders including TomPhilp (CEO, Maximum Information), Elisabeth Viktor (Severe Convective Storm Lead at Swiss Re), andGuido Felder (Climate Risk Analytics Lead at Zurich) to address one of the industry's most pressingchallenges. The panel emphasised that model validation within the Oasis LMF goes beyond technicalverification - it ensures catastrophe models accurately represent risk for users' specific portfolios whilemeeting regulatory requirements and maintaining reliability under varying conditions. Central to thediscussion was quantifying uncertainty across the entire modelling chain, from hazard representation toloss calculation. The session highlighted how users can be better supported in model evaluationprocesses, addressing critical questions around exposure assumptions, vulnerability assessments, andloss output validation. This approach reflects the industry's urgent need for standardised evaluationpractices, driven by the reality that model outputs can vary dramatically for the same risks, making robustvalidation essential for regulators, internal governance bodies, and pricing actuaries.
Harnessing Innovation in Catastrophe Modelling: A Journey to OpenSolutions
SCOR's Global Head of Cat R&D, Simos Koumoutsaris outlined the reinsurer's future vision forcatastrophe modelling, highlighting their strategic focus on developing specialised models wheretraditional vendors fall short, exemplified by their in-house California Wildfire model. SCOR's roadmapemphasises transitioning from manual, resource-intensive workflows to automated API-driven processesthrough a Central Exposure Repository and Model Execution tools. Their proposed OASIS Model Builderrepresents a significant leap forward, enabling users to input hazard footprints and vulnerabilityparameters to automatically construct complete catastrophe models, dramatically reducing developmentcomplexity and time-to-market for innovative risk assessment solutions.
Adam Podlaha (CEO, Aon Impact Forecasting) shared how the organisation engages across multipleroles within the Oasis ecosystem—from model development and deployment to shaping communitystandards. A long-standing contributor to Oasis, IF played a key role in the development of dynamicfootprint capabilities and continues to support the evolution of the Oasis model definition format andOpen Data Standards (ODS) for property characteristics. The presentation highlighted a growingsophistication in model deployment strategies. Impact Forecasting now delivers models via theirproprietary ELEMENTS platform as well as in Oasis-compatible formats, expanding accessibility acrossthe market. By embracing cloud-native architectures and standardised APIs, Aon is driving scalability,performance, and smoother integration across diverse platforms.
James Lay (AVP Product Management, Verisk) discussed the newly renamed, ' Model Exchange', nowpart of Verisk. Via the platform clients have access to a vibrant community of model vendors via a singleplatform. The platform will continue to be developed as a standalone platform where numbers of modelsand model vendors available by the platform are growing. The model exchange team and Verisk are alsofocusing their efforts on working on a robust method to convert data between OED and CEDE exposureformats.
Peter Michailow (Senior Risk Manager, Baloise Group) presented their remarkable five-year journey(2019-2024) from regulatory challenge to successful implementation of a state-of-the-art earthquakemodel for Switzerland. Baloise partnered with GEM for hazard modelling and Guy Carpenter forvalidation, creating a hybrid technical architecture using OpenQuake for complex hazard processing andOasis for streamlined daily operations. Their success story demonstrates how collaborative innovationcan conquer regulatory demands while maintaining strategic sovereignty through opensourceframeworks.
Tom Hughes (Senior Exposure Manager, Aviva) shared how Aviva's decision to use Oasis models wasmade remarkably straightforward through their partnership approach. The efficiency of working with aunified team—formerly known as the NASDAQ group—eliminated complexity by offering a single contractstructure. This streamlined approach made it an easy decision for Aviva, demonstrating how simplifiedvendor relationships can accelerate Oasis adoption.
Global Head of Cat Risk Management, Markus Stowasser emphasised Allianz's strong support for theOasis community, highlighting their strategic belief that standardisation delivers real value in data,catastrophe frameworks, and financial modelling. This approach fosters diverse perspectives on riskassessment while building robust catastrophe risk management capabilities. Allianz has backed up thisphilosophy with substantial investments in proprietary tools and frameworks, creatingvendorindependent, highly automated systems integrated into their core insurance platforms.
Looking Ahead: The Future of Catastrophe Modelling
The final panel brought together Balz Grollimund (Head of Catastrophe Perils, Swiss Re) , Rick Thomas(Managing Director, Acrisure Re), Paul Della-Marta, and Dickie Whitaker for a thought-provokingdiscussion about the evolving needs of underwriters and portfolio managers. They challenged us to thinkbeyond incremental improvements and imagine truly transformational changes to underwriting workflows.
The conversation explored several questions: What will the future underwriter actually need frommodelling tools? How do we balance AI/ML automation with human judgment? What's the optimal levelof transparency for commercial catastrophe models? The panellists emphasised that we're not looking toreplace human expertise but rather to enhance it, creating more efficient exposure management toolswhile preserving the critical thinking that makes great underwriters invaluable.
What emerged was a vision of empowered professionals equipped with sophisticated, automated toolsthat handle routine tasks while freeing them to focus on complex risk decisions that require human insightand experience.
As the conference concluded, several key trends emerged as priorities for the industry's futuredevelopment. Multi-peril integration is becoming essential as organisations seek comprehensive
coverage across various natural hazards. Climate change adaptation requires enhanced modelling fornon-stationary conditions, while automated workflows promise to reduce manual intervention andimprove efficiency. Advanced analytics and AI/ML applications are moving from experimental tooperational status, offering new capabilities for risk assessment and model validation.
Perhaps most significantly, the conference demonstrated that the industry's evolution towardcollaborative, open-source solutions is accelerating rather than slowing. The Oasis community'scontinued growth, the expansion of standardised data formats, and the increasing participation of majorindustry players signal a fundamental shift in how catastrophe modelling will be conducted in the yearsahead.
Stay Connected
If you're already part of our community, thank you for being part of this journey. The innovationsshowcased in Zurich are possible because of contributions from people like you.
If you're considering joining us, there's never been a better time. With 100+ models available, expandingstandards, and a growing community of experts, the Oasis ecosystem offers unprecedented opportunitiesfor innovation and collaboration.
Missed the conference but want to dive deeper? Visit the website or resources, documentation, andinformation about upcoming events. at : Oasis Loss Modelling Framework | Open sourcecatastrophe modelling platform
The next chapter of catastrophe modelling is being written by our community, and we'd love to have yoube part of that story.
Want to contribute to future newsletters or share your Oasis success story? Contact us please reach outat harriet.smith@oasislmf.org. We love highlighting the work our members are doing!