La Boétie
Research build · two engines

AssurQuant.

Quantitative and quantum engineering for the numbers insurers bet the company on.

Two engines for insurance risk. One prices what a company should pay for its cover. The other runs the Solvency II capital calculation in full, twenty four billion valuations, and prices it on a real quantum machine. The quantitative engineering we take on for clients.

assurquant · solvency ii capital
the calculation24 billion valuations
runexact, in full
priced ona real quantum machine
at stake€99M a year, per point

The largest routine calculation in insurance, run in full and priced on real quantum hardware.

Solvency II in full · quantum amplitude estimation · real hardware

What AssurQuant is

Two engines for the numbers insurers bet the company on.

AssurQuant prices insurance risk at both ends. One engine tells a company what its group health cover should cost and where it is bleeding money. The other runs the Solvency II capital calculation, the largest routine computation in finance, in full, and prices the same risk on a real quantum machine.

01

Prices a company's group health cover and pinpoints where it overpays the market.

02

Runs the Solvency II capital calculation in full, twenty four billion valuations, not the approximation the industry settles for.

03

Prices the same risk with quantum amplitude estimation on a real quantum machine, not a simulator.

How it is built

Three engines, one risk, priced end to end.

A pricing model for what a company pays, a Monte Carlo engine for what an insurer must hold, and a quantum kernel that prices the same risk on hardware the rest of the field only writes about.

The pricing engine

What a company should pay, against the market

It places a company's group health cover against the whole market and returns a number with a verdict: overpaying, in line, or under. Where the real premium is on record, it shows that figure outright.

The capital engine

The whole Solvency II calculation, in full

The capital an insurer must hold comes from nested Monte Carlo: thousands of scenarios, each demanding thousands of revaluations, twenty four billion computations for one number. Most of the industry runs an approximation. This runs it exactly, on a closed form validated model, and benchmarks the fast path against the exact answer.

The quantum kernel

Amplitude estimation on a real machine

The same capital and risk measures, expected loss, VaR, CVaR and the tail, priced by iterative quantum amplitude estimation on a real quantum computer with error suppression. Not a simulation of quantum. Quantum.

The scale

The size of the calculation, the money on it, and the machine.

Getting the capital number slightly wrong is not an academic problem. It is a nine figure one, every year, across the market.

€99Ma year, per point of error

What a single point of Solvency II capital error costs the French life market annually.

24 billionvaluations, one capital run

Every scenario in the nested capital calculation, computed in full rather than approximated.

0.8%on real quantum hardware

How close the quantum amplitude estimation lands to the exact value, on a real quantum machine.

What comes next

Built for the machine that is coming.

Both engines run today, the classical one exact, the quantum one on real hardware. The frontier is the crossover, the point where the quantum kernel overtakes the classical run, and AssurQuant is built to cross it the day the hardware allows.

Live

Both engines, live

Pricing and capital both run today, the capital number computed classically and on a real quantum machine.

Research

The crossover

Pinning the exact point where quantum amplitude estimation overtakes the classical run on real hardware.

Planned

Peer reviewed

The paper is written, with reproducible figures. Submission to actuarial and quantitative venues is next.

What this means for you

Quant and quantum engineering for the numbers you cannot get wrong.

Heavy quantitative and quantum computing, aimed at risk that regulators and boards act on. When your numbers have to be right, this is who builds them.

Quantum computing

Real algorithms on real quantum hardware, error suppression, and a hard read on exactly where quantum overtakes classical.

Large scale simulation

Nested Monte Carlo at tens of billions of valuations, validated stochastic models, variance reduction, and GPU acceleration built for it.

Data engineering at scale

Tens of thousands of documents turned into clean, structured training data by pipelines that resume, dedupe and never lose a field.

Applied ML that holds up

Models with calibrated intervals, measured out of sample, that have to beat a real benchmark before they ship.

Regulated domain modelling

Solvency II, actuarial calibration, and the figures insurers answer to a regulator for, built to a standard you can put in front of one.

Reproducible from clone

Pipelines with test and regression gates, and results anyone can rebuild from a clean checkout.

FAQ

Questions

Two quantitative engines for insurance: one prices what a company should pay for its group health cover, the other runs the Solvency II capital calculation in full and prices the same risk on a real quantum machine. It is a La Boetie lab build, and it shows the quant and quantum engineering we do for clients.

For the numbers you cannot get wrong.

Quantitative modelling, quantum computing, simulation at scale, wherever a wrong figure is expensive. Tell us the number you need to trust, and we will build the engine that earns it.