(Re)insurance decision layer on Palantir Foundry & AIP

Commercial insurers and reinsurers do not suffer from a lack of systems. They suffer from the gaps between them. Palantir Foundry and AIP close those gaps, turning every function into one decision layer.

Why now

The bottleneck moved from model to context

Insurance leaders are moving AI into underwriting, claims, distribution, and service. The model can answer; the real question is whether it has the governed context to answer safely:

Policy wordingClaims historyExposurePricing assumptionsBroker economicsReservesAuthority limits
01

AI moved into the business

Boards now expect AI in core workflows, not innovation demos.

02

The decisive context is unstructured

Broker emails, loss runs, adjuster notes, PDFs: the context lives where reporting never modeled it.

03

Governance is part of the product

It must respect permissions, authority limits, and lineage from the first workflow.

Insurance doesn't lack systems. It lacks one connected model

Conventional landscape

  • ×Each function owns its own applications and definitions of truth
  • ×Decisions that cross boundaries require manual assembly
  • ×AI pilots sit beside workflows instead of changing the decision
  • ×Finance sees results after the operational window has closed

The Palantir play

  • Submissions, policies, claims, and capital become connected objects
  • AIP reasons with permissioning, lineage, and authority limits
  • Underwriters price with full portfolio and claims context
  • Finance steers from live signal, not stale committee packs

Palantir makes the insurance value chain computable. 10x makes it real in production.

The play

One ontology connecting policies, claims, exposure and capital

Palantir FoundryOntology layer

One live model your teams and AI both work from.
Every answer is grounded in your real data and traces back to its source.

One model across every system

Policy & treaty
Wording, limits, layers
Claims
Loss runs & reserves
Pricing
Technical, experience, exposure
Distribution
Brokers, submissions, cedants
Exposure
Location, peril, accumulation
Finance
Premium, capital, IFRS 17
01

One connected model

Every function works from the same live data, not its own copy.

02

Decisions, not dashboards

Teams act on connected policies, claims, and capital directly in the workflow.

03

AI grounded in truth

Every answer is backed by source data, with a human in the decision.

Your systems of record stay. Foundry becomes the decision layer across them.

Palantir Foundry connects the value chain

Underwriting to finance, on one model

Select a function to see the use cases, pain, and connected data inside that part of the organization.

Selected function

Underwriting & Pricing

Turn submissions, appetite, pricing, and policy or treaty history into one underwriting workspace.

Submission triaging

01

Pain

Underwriters lose time assembling the risk picture before they can decide what deserves attention.

Foundry connects

submission datapolicy and treaty historypricing databroker and cedant dataappetite rules

Focus underwriting effort on the business most likely to fit appetite and move portfolio goals.

Treaty & account pricing

02

Pain

Loss history, exposure, and assumptions live in separate files, so the view is stale before it is priced.

Foundry connects

loss historyexposure datapricing modelspolicy and treaty termsactuarial assumptions

See the technical price with every driver traceable back to source data.

Rate adequacy & discipline

03

Pain

Rate erosion and discretionary credits accumulate quietly until discipline has already drifted across the book.

Foundry connects

pricing datapricing parametersunderwriting actionssubmission dataportfolio plan

Catch directional drift and act before the book is already written.

Worked example

A live portfolio question, answered from connected data

An underwriting executive asks why profitability is drifting in a property segment, and whether recent claims should change pricing.

Can the team answer live, without waiting for a manual pack?

Connected objects

SubmissionBroker docs, intake, appetite
Policy & treatyLimits, wording, layers
PricingTechnical, experience, exposure
ClaimsLoss runs, severity, leakage
ExposureLocation, peril, accumulation
Broker & cedantBook, relationship, economics
FinancePremium, loss ratio, plan
CapitalAppetite, ROE, cessions

Reference demo: AIP Analyst answering an insurance operating question over connected business objects rather than a static report.

Before

Teams assemble a committee pack from spreadsheets, extracts, dashboards, and model outputs.

With Foundry + AIP

AIP Analyst queries the ontology, explains drivers, and links the answer back to the operating objects behind it.

Selected case studies

Three places we have already built the pattern

Not a capability catalogue. Three examples showing where the decision layer has already moved from slideware to production work.

Why 10x Partners

Palantir gives you the platform.
10x turns it into a production operating layer

100+
Foundry use cases across our careers
4y
Average platform experience
Weeks
To production, not quarters
  • /Founded by ex-Palantir forward-deployed engineers and ex-Insurance and Reinsurance leads.
  • /Senior-only. No juniors, no handoffs, no learning curve on your budget.
  • /Official Palantir partner. We build production operating systems, not slideware.
Meet the team behind your (Re)insurance build

Frequently Asked Questions

Explore other industries

Build the (Re)insurance decision layer.

We help (Re)insurance teams turn Palantir Foundry & AIP into production systems, not another dashboard.

Get in touch