How We Invest
Twelve years of insurance operating experience, applied at the seed stage, in the European market that needs it most.
The Underwriting Model Is Being Replaced
Insurance has operated on the same mathematical infrastructure for a century. Actuaries pool risks, estimate future losses from historical loss tables, and price policies accordingly. The model is directionally sound but structurally backward-looking — it answers the question "what did similar risks cost in the past?" rather than "what does this specific risk cost right now?"
Machine learning changes the calculus. Models trained on high-frequency data — telematics, satellite imagery, transaction flows, medical records with consent — can assess individual risk continuously. Not at policy inception from aggregate loss curves, but at every moment of the insured period. The shift from pooled-loss-table pricing to continuous individual risk assessment is not incremental improvement. It is a structural replacement of the underwriting methodology that has governed European insurance for a generation.
Continuous repricing at the individual contract level was not mathematically tractable before. ML-based underwriting doesn't just reduce fraud or automate claims handling — it makes possible an entirely new class of insurance product that incumbents cannot manufacture using their existing actuarial infrastructure. That is the opportunity we are backing.
Why Seed Stage
We invest at seed because the foundational architectural decisions are made in the first eighteen months and are expensive to undo. The data pipeline design determines which signals a model can learn from — and redesigning it at a later stage means rebuilding the product from beneath. The carrier partnership strategy determines whether a startup functions as an MGA, a technology vendor, or a fully licensed insurer — and each path has different regulatory capital requirements and distribution economics that compound over time. The regulatory positioning, particularly under Solvency II and EIOPA AI governance guidelines, shapes the firm's operating model from the first product decision.
Founders making these decisions at the seed stage rarely have access to operators who have sat on both sides of the table — as actuaries pricing the risk and as product leaders designing against it. That is precisely what Morildsen brings to these conversations. Not as passive advisors, but as active thinking partners at the moment the architectural choices get locked in.
Why Europe
The European insurance market is structurally underserved by deep InsurTech capital. US-centric venture firms consistently underestimate the regulatory complexity: Solvency II capital requirements, EIOPA AI governance guidance, GDPR interaction with underwriting data, and a carrier landscape distributed across 27 national markets with distinct regulatory regimes. These are not obstacles that dissolve with product-market fit. They are the operating environment. Founders who have read the Solvency II directives and understand EIOPA's expectations for model explainability have a structural moat that US entrants consistently underestimate.
Morildsen was built for this market. Our LP base includes Nordic institutional investors with deep insurance sector experience. Our networks reach Chief Underwriting Officers and Heads of Claims at European tier-1 carriers. We understand the Solvency II balance sheet implications that most seed funds treat as a legal footnote — because we have worked inside them.
Check Size and Structure
Initial checks of €1.5M–€3.5M at the seed stage. We reserve meaningful capacity for follow-on, typically targeting 2× the initial check in the first follow-on round. We back startups with the technical depth and founder conviction to build durable ML-driven insurance infrastructure — not feature layers bolted onto legacy policy administration systems. The founders we back understand both what their training data is saying and where the model's blind spots lie.
What We Look For
Technical Depth
Founding teams who understand both the insurance domain and ML system design — not adding a predictive layer onto a process that is still fundamentally manual. We look for teams who can reason about model drift in claims frequency data, feature engineering in adversarial fraud environments, and the EIOPA explainability requirements that create real architectural constraints. The standard is high because the problems that surface later are proportional to the depth of understanding at the start.
Distribution Insight
Teams who have thought carefully about how insurance products reach policyholders — whether via MGA structures with carrier fronting arrangements, embedded API distribution at the point of sale, or direct digital acquisition with full carrier licensing. Distribution in insurance is not a go-to-market question. It is a regulatory capital question, a reinsurance treaty question, and a question of who owns the customer relationship under the Solvency II balance sheet. Founders who treat it as a launch decision rather than a structural one consistently encounter the same walls.
Regulatory Fluency
European insurance regulation is not an obstacle for the right founder — it is a moat. Solvency II capital requirements, EIOPA AI governance guidance, and GDPR constraints on underwriting data are entry barriers that protect well-positioned startups from fast-following incumbents. We back founders who have read the directives, not just the summaries. We do not invest in InsurTech startups that intend to address European markets from a US regulatory posture — the two operating environments are structurally different, and experience in one does not transfer cleanly to the other.
How We Work with Founders
Frederik's five years advising Nordic insurers on digital transformation gives Morildsen direct access to the decision-makers who actually sign vendor contracts at tier-1 European carriers — Chief Underwriting Officers, Heads of Claims, and digital transformation leads who are actively evaluating new partnerships. These are not generic introductions. They are warm introductions to named contacts at specific carriers, based on a working understanding of which insurers are piloting in which lines and who owns the internal mandate. Portfolio companies working on their first carrier partnership have used this network to compress what would otherwise be a nine-month procurement cycle to under four months. The reduction comes from starting the conversation with the right framing in the right room, not from following up faster.
Astrid holds monthly one-to-one sessions with founding CTOs navigating the specific engineering challenges of ML models in production insurance environments. These challenges do not appear in general ML literature: claims frequency data drift during exogenous events like pandemic periods, EIOPA explainability requirements that create direct tension with ensemble model architectures, adversarial feature engineering in fraud detection where policyholders adapt to the scoring signals over time. Astrid built and operated fraud detection processing €2B daily transaction volume at a Nordic banking group — in a regulated environment where model decisions had direct legal standing. She brings that production-scale, regulated-environment experience directly to portfolio architecture reviews, not general software engineering advice.
Founders do not navigate later-stage fundraising alone. We actively coordinate introductions to funds with demonstrated appetite for European InsurTech at the growth stage, including funds where Morildsen LPs have co-invested and where we can speak to specific partners' investment patterns. When we make an introduction, we provide context on what that investor has underwritten before and what they typically need to see in the diligence process — which means founders can prepare specifically rather than generically. The goal is to have three or four credible conversations in motion before the process formally opens, so founders are negotiating from a position of genuine optionality.
Mads takes a board observer or board seat in the majority of Morildsen investments, with a focus on governance specific to regulated financial entities. The issues that come up are not generic startup operational questions. They include audit committee structure for entities operating under insurance intermediary authorisation, managing the FCA or DNB relationship during periods of rapid product iteration, and the specific governance obligations that attach to holding an insurance distribution licence. The actuarial background is directly relevant here: twelve years pricing and managing treaty risk at a Scandinavian reinsurer means Mads has worked within the same regulatory frameworks that portfolio companies are building under — not just read about them.
Fund Scale
Total assets under management: $95M across 14 portfolio companies.