HomeSolutions — Insurance & warranty
Insurers, warranty providers and lenders

A claim is won on what you can show. Not on what you believed at the time.

When an expensive asset fails, the argument is almost never about whether it failed. It is about what it had been through, who knew, and when. That argument is settled by paperwork — and most operators discover what their paperwork is worth only after they need it.


Where the money actually goes

The failure costs you once. The dispute costs you again.

Three separate bills follow a major asset failure, and only the first one is the equipment.

The replacement

The part, and the capacity that earns nothing while you wait for it. On large power equipment that wait is now measured in years, not weeks.

The bill everyone expects

The denied claim

Warranties carry operating conditions — temperature, cycling, depth of discharge, load. When a claim is challenged, the question is whether the asset was run inside them. If you cannot show it, you paid for coverage you cannot collect.

The bill nobody budgets for

The re-rated premium

Insurers and lenders reprice on what they can verify. A fleet with no defensible history is priced as an unknown, and unknowns are expensive — every year, whether or not anything ever fails.

The bill that never stops

Why a forecast is not enough

Nobody underwrites an opinion.

A number on a dashboard tells an insurer nothing they can act on.

Put yourself on their side of the table. Someone hands you a score — 74, amber, four months — and asks you to price risk against it. You cannot. You do not know what produced it, what it assumed, what data it was missing, or whether it would have said the same thing last quarter. There is nothing there to check, so there is nothing there to trust.

Now hand them a record instead: the call, the date, what it was based on, which figures were measured and which were assumed, and what the assumptions were worth if they moved. That is a document someone can examine, disagree with, and price. The disagreement is the point — you cannot argue with a score, and an argument you can have is an argument you can win.

This is the whole reason we changed what we hand you. The forecast was never the product. The record is.

Three parties, one requirement.

Your insurer needs to know what the equipment experienced and what you did about it. Evidence of a condition you knew about and ignored is worse than no evidence. Evidence of a condition you found early and acted on is the case for a better rate.

Your OEM needs to see the operating conditions when a warranty claim is disputed. Both sides are arguing about the same history. Only one of you usually has it written down.

Your lender or offtaker needs the asset to still be worth something at the end of the term. A fleet with a maintained record is a different asset, on paper, from an identical fleet without one.


What a record has to contain

Enough that a stranger could check it.

That is the whole test. Not whether the call was right — whether someone who was not in the room can follow how it was reached.

The call, with a date on it. What we said would happen, when we said it, and what we recommended doing. A recommendation that appears only after the failure is not evidence of anything.

Dated

What it was based on. Which measurements, over what period, from which equipment. A claim assessor can ask where a number came from and get an answer.

Traceable

Which figures were assumed. Assumptions marked as assumptions, with what they were worth if they moved. Every model rests on some — the difference is whether yours are visible.

Labelled

When it declined to answer. Where the inputs could not support a call, the record says so and says what was missing, rather than producing a number anyway.

Stated

That last one does more work than it looks. A system that always produces an answer cannot tell anyone which of its answers were worth having. One that declines when it cannot see enough is telling you exactly where its own evidence runs out — which is the first thing anyone assessing a file wants to know. How the abstention gate works.


Where this already exists

It is not a separate product.

The record is what the platform produces. Two places to see the shape of it.

UptimePowr

Every recommendation is written as a justification record: the call, the basis, and the cost assumptions labelled as assumptions. Built for someone in finance to read, not only someone in engineering.

Battery fleets

Warranty curves rest on reference conditions your cells never actually see. We tell you which parts of that number to trust, how far, and where it runs out — which is the part a warranty conversation turns on.


● Read this before you quote us

What we have not proven.

We would rather you hear this from us than from your broker. The argument on this page is that the record contains what a file needs. That is a claim about the document. It is not a claim that anyone has yet accepted it, and we are not going to blur the two.

Talk to us with the limits open →
Not yet true
No insurer has priced a policy on this record. We have no carrier relationship and no underwriting partnership.
We have no premium-reduction figures, because we have never run one through. Anyone quoting you a percentage is guessing.
We have never taken a warranty claim to resolution. Whether your OEM accepts the record is between you and them.
Choir has no deployed customer reference. Everything published so far runs on public or simulated data.
We are not brokers, underwriters or advisers, and nothing here is insurance advice.

Find out what your paperwork is actually worth.

A read-only assessment runs on the data you already have and produces the record in its finished form. Show it to your broker, your OEM or your board and let them tell you what it is worth.

Software-first · no raw data leaves your network · every number traces to a dated report