Get Your Accuracy Number
For insurance brokers & insurers

Insurer post that files itself. To the right client, first time.

We measure it on 500 of your own incoming documents in 14 days: what got matched, what got misfiled, and what a human still has to see. Before you commit to the build.

Fixed pilot fee. Below the matching threshold we agree together, you pay nothing. No per-document pricing, no annual licence.

What sorting the post really costs

Nobody budgets for filing. Everybody pays for it.

A broker with fifteen insurer relationships receives fifteen document conventions. Somebody opens each one, works out whose it is, and drags it to a folder. Count how many times that happened last month.

60–120 sec

per document to open it, identify the client and policy, rename it and file it. Multiply by everything the insurers sent you.

2–5%

of documents filed against the wrong client or the wrong policy year. You find them when someone goes looking during a claim.

1 day

typical lag between an endorsement arriving and it being visible on the client record, which is a day your account handler answers from memory.

1 person

who recognises each insurer's layout on sight, and whose absence backs the post room up within a week.

Do the math with your numbers

Three fields. This is what filing insurer post costs you per year.

Per month€3,733
Per year€44,800
The pilot that ends it€4,000 once

Excludes the cost of a misfiled document surfacing during a claim, which is the expensive kind and does not show up in a time study.

The mechanism

How a document finds its own client record

Filing is a matching problem, not a reading problem. The hard part is not pulling a policy number off a page. It is deciding which of your four Smith Ltd records the page belongs to, and admitting when it cannot tell.

Classify the document

Schedule, endorsement, renewal invitation, cover note, premium advice, claims correspondence. Each type carries different identifiers and different urgency, so the type is decided before anything is extracted.

Match against your book

Policy number, insurer reference, insured name and risk address are all read, then scored together against your client and policy records. Agreement across several identifiers is what produces confidence, not any single field.

File or escalate

A confident match files itself with the document type and date recorded. An ambiguous one goes to your team with the candidate records side by side and the reason for the doubt named.

What this looks like on a real insurer document

An insurer endorsement on the left. What the layer extracts and decides on the right. Illustrative example with fictional data.

Policy Servicing · Leeds LS1 4AP
Endorsement NG/EN/88214
Effective: 01.08.2026
Policy: NG-CP-2025-77310
ItemChangeWasNow
1Buildings sum insured, Unit 41,850,0002,140,000
2Business interruption indemnity12 months24 months
3Additional premium, pro rata-1,284.60
Additional premium 1,284.60
IPT 12% 154.15
Total due 1,438.75
Insured: Hartley Logistics (Northern) Ltd · agent ref not quoted on this document
Extracted & matched
Document typeEndorsement not a renewal invitation99.4%
InsurerNorthgate Underwriting99.8%
Policy numberNG-CP-2025-7731099.7%
Effective date01.08.202699.5%
Sum insured change1,850,000 → 2,140,00098.8%
Additional premium1,438.75 reconciles with IPT line99.1%
Client recordTwo Hartley Logistics entities share this policy number history61%
⚠ The document is read correctly but the client match is not certain → it goes to your team with both candidate records shown and the conflict named, rather than filing against the likelier one.
Filing record created
Endorsement · NG-CP-2025-77310 · eff. 01.08.2026 · AP 1,438.75 · awaiting client confirmation

The offer

The Broker Post Intake Pilot

One fixed fee. Fourteen days. Your real inbound post. A matching accuracy number at the end that tells you, and us, whether the full build is worth doing.

Not a demonstration on tidy sample documents. The pilot ends with your own month of post processed and a per-insurer breakdown you can argue with.

  • 500 of your real inbound documents processedEvery insurer you deal with, including the two that still send scanned paper.
  • Matching accuracy report, per insurer and per document typeBecause a renewal invitation from one insurer and an endorsement from another fail for completely different reasons.
  • Misfile analysisWhich documents would have gone to the wrong record, and what the identifying information was missing.
  • Working filing path into your broker systemA test batch filed end to end, so you see the integration behave against your own client data rather than a sample set.
  • Fixed-price quote for the full buildScoped from what we measured on your post. Valid 90 days.
  • ROI calculation on your own volumesYour document count, your handler cost, your misfile rate. If the arithmetic does not work, you will read that in the report.
€4,000 fixed

Against a problem costing most clients €40,000–70,000 a year.
Credited in full against the build if you proceed within 90 days.

Get Your Accuracy Number

Your side of the pilot: about two hours, total

45 minkickoff call to agree the accuracy threshold and pick the sample
10 minforward an email folder or grant read access to the document inbox
60 minresults session: the accuracy report, the exceptions, the quote

No software to install, no workshop series, no project team on your side. Your people keep working; we work on copies.

The Accuracy-or-Free Guarantee

Before we start, we agree the accuracy threshold that makes this worth deploying. Together, in the engagement letter. If the measured result on your documents comes in below it, the pilot is free. We can offer this because we scope honestly: if we don't think your document mix can hit the number, we decline the pilot and tell you why.

Pilots run one at a time per engineer, so they start in the order requests arrive. Current lead time: about two weeks

Your actual alternatives

What else you could do with this problem

Do nothingHire another handlerVertical SaaS productBraynex pilot → build
Matches to your own client book✕ manual every time✓ if they know the book✕ matches on policy number alone✓ scored across several identifiers
Cost over 3 yearsrising with your book€110–160k+per-document pricing, grows with volumeknown, fixed, quoted
Proof before commitment-✕ demo on their documents✓ measured on your post
Admits when it is unsure-✓ usually asks✕ files its best guess✓ escalates with the reason
Survives an insurer changing its layout✓ after retraining the person✕ template breaks✓ trained on your mix, retrained on ours

Why believe us

We build document systems for businesses that cannot afford to be wrong

Medical / Legal

MedChrono

AI that converts medical records into legally defensible timelines. Accuracy-critical, regulated, every extraction traceable to its source page.

Financial advisory

Advisory documentation

Client meeting recordings become compliant advisory documents automatically, adopted across the practice for every client meeting.

Vision AI at volume

Sightline

Production vision-AI moderation for marketplace listings: real ML in production at volume, not an API wrapper.

Swiss entity (Zug) · EU-region hosting · GDPR & revised FADP compliant · named senior engineers you meet before signing

Fair questions

The things you should be skeptical about

Our broker system already has document management. Why isn't that enough?

It stores documents well. It does not decide which client a PDF belongs to, which is the part your team actually spends time on. We sit in front of it: by the time a document reaches your document management, it already knows what it is and whose it is.

What happens when two clients have similar names?

That is the case the pilot is designed to measure, not to hide. Where several identifiers agree, the document files itself. Where the name matches but the policy number history does not, it stops and shows your handler both candidates. The report tells you how often that happened and why, so you can decide whether the residual manual rate is acceptable.

Can this write into our broker management system?

That is the part we prove first, in a test environment, during the pilot. Some systems expose a clean interface and some need a supported integration path; if yours turns out to be the hard kind, you learn that in week one for a fixed fee rather than in month four of a build.

What if we'd rather buy one of the broking-specific products?

Sometimes you should, and the pilot report will say so plainly. If your post is dominated by a handful of insurers who all send structured data, an off-the-shelf product may cover you and we will tell you that rather than sell you a build. You will have paid a fixed fee for a clear answer either way.

Where does our data go?

EU-region hosting, a GDPR-compliant processing agreement, and a contract with our Swiss entity in Zug. Nothing trains a public model. Your own compliance lead can review the setup before the pilot starts, and the pilot sample can be narrowed to a subset you are comfortable sharing.

Why should a company our size get your senior engineers?

Because we are structured for exactly this engagement size. The large consultancies staff €60k projects with juniors or decline them; we deliver them with the people you actually meet. That is the trade we have built the company on.

Start here

Claim a pilot slot

Tell us roughly how much post you receive and which systems it has to land in. You get a reply from an engineer within one business day, and a straight answer about whether this is worth piloting on your book.

FS

Florian Shabani

Lead Engineer · runs your pilot

The people on this page are the people on your project. No handover to a delivery team you never met: the engineer who scopes your pilot is the engineer who builds it.

One pilot at a time per engineer · current lead time: ~2 weeks

No newsletter, no drip sequence. One reply from an engineer, within one business day.