Each tool turns a research question into something you can handle: change an assumption, watch the result move. They cover cyber pricing, foundation-model accumulation, climate insurability, EU AI Act compliance and value-chain dependency. No account, no payment, no data collected.
Cognitive marginal cost simulator (Human vs Cloud API vs Local Architecture). Includes cyber premiums, AI tax regulation and the Make vs Buy tipping point per Coase.
Actuarial engine modeling the systemic exposure of LLMs. Includes the EU AI Act (4 tiers), mitigation protocols (RAG, Red-Teaming), Quota Share structuring and ESG scoring.
27 bilingual actuarial converters: loss ratio, combined ratio, burning cost, lognormal VaR, chain ladder, Bornhuetter-Ferguson and more.
Browse and compare Lloyd's cyber clauses side by side (LMA 5564–5567, Property, Marine, Reinsurance). Official texts, pedagogical translations, write-back analysis.
AI data center insurability screen seen through a sustainable reinsurer. Places the site on SCOR's strategy × operations matrix, computes an accumulation index, fires ten business rules and exports a PDF underwriting memo.
Quantify the logical concentration of your AI book by foundation model (GPT-4o, Gemini, Claude...). Computes an accumulation index, Model PML and EU AI Act flags.
Frequency-severity actuarial model simulated in the browser: annual loss distribution, return-period curve (99.5% VaR) and insurance-program sizing (retention, limit, premium). Parameters and sources shown.
Projects an exposed property's average annual loss from present to +4 °C, against market insurability thresholds and France's CATNAT scheme (surcharge, CCR, state guarantee). The tipping point, parameters and sources shown.
A diagnosis of an AI system against the EU Regulation: risk tier, obligations dated by role, liability exposure and the insurance covers that respond. Sources shown, neither legal advice nor certification.
A map of the AI value chain (compute, cloud, model, orchestration, application): locate single points of failure, measure the cascade and the systemic exposure of a dependency. Parameters and sources shown.
Automated intrusion testing on an insured's LLM to certify E&O exposure before reinsurance placement. On-premise architecture, zero external API calls.
Outputs are orders of magnitude meant to frame a discussion, not firm prices or inventory figures.
Every parameter is exposed and editable. Nothing is computed behind an opaque screen: disagreement over an assumption is settled by changing it.
Same parameter set, same result. PDF exports carry the assumptions used, so the calculation can be rerun or challenged.
These are research instruments. They are not insurance advice, not an offer, and not a certified pricing engine.
The data behind these tools comes from a documented directory: scope, methodology, intended use and limits of each source, with its links to our articles and glossary terms.