Specialty insurance keeps solving the same problem in fifty places. Noor solves it once.

One layer over the systems you already run: every system addressable, one record of the business underneath, and the workflows, endpoints and agents built on top. Bordereaux, broker transacting, submissions, reporting — same layer, no re-platforming.


The shift

It looks like fifty problems. It's one problem, fifty times.

Follow a single risk from the moment an insured sends an application form to the moment the money lands. Each of these gets treated as its own project, with its own integration, its own spreadsheet and its own person holding it together.

  1. 01 · Submission — The risk that arrives as an email with attachments, and gets typed into a system by hand.

  2. 02 · Bind — The bind that needed a referral, and got one because somebody remembered what the binder allows.

  3. 03 · Mid-term — The survey that found a hazard in March, was filed as a PDF, and never reached the record.

  4. 04 · Claim — The claim that takes months to get visibility of.

  5. 05 · Month-end — The bordereau that takes a week to assemble and a month to reconcile.

  6. 06 · Exposure — The exposure data that arrives in hundreds of columns and still doesn't tell the carrier what it is carrying.

  7. 07 · Settlement — The premium sitting still because a field describing a building is in dispute.

And behind every one of them, the same cost paid again and again: the new carrier, the new product, the new system — each one a project from scratch.


What Noor is

Noor is the digital wrapper around the systems this market already runs.

Nothing inside the wrapper changes. The policy admin system, the rating engine, the document generation, the claims process and the finance stack all keep doing exactly what they do today. Noor goes around them and does two things: on the outside it removes the friction of doing business with everyone else in the value chain, and on the inside it gives each business the tools to optimise how its own work gets done. Same systems. Both sides.

  1. Outward · Doing business with the value chain

    Every relationship in this market costs both sides something: a portal to log into, a template to fill, a file to assemble, a format to match, an integration to build. Noor turns each of those into a connection made once.

    MGAs spend their time underwriting, rather than working out how to connect to each new broker or reconciling data on somebody else's behalf. Brokers get an instant, personalised quote back where they already work — no portal, no re-keying — with the client pack and the market comparison already built. Carriers pull the data they need, in the shape and at the cadence they specify. Reinsurers and capital providers see exposure as it changes rather than as it is reassembled. And when the request arrives from software rather than a person — a broker's AI, a client's AI — it gets the same answers, from the same endpoints, under the same rules.

    The relationships stay where they are. The friction does not.

  2. Inward · Running the operation inside

    The same connections and the same record power the work inside each business. An MGA can read and structure a submission the day it lands, run workflows across systems that were never designed to talk to each other, raise exceptions to the right person instead of finding them at month-end, and put AI to work inside its stack rather than beside it.

    None of it requires a migration, a re-platform or a change to how the underwriting gets done. The people in the middle stop being the integration layer.

Noor is not a policy administration system, not a data warehouse, and not a standard anyone has to adopt. It is the layer that makes the systems already in place work together — without changing a single one of them.


The platform

What's inside the wrapper.

Layer 01 · Connections

Every system addressable, whether or not it was built to be. Read and write across policy admin, rating, documents, claims and finance, and expose any capability outward as an API or MCP endpoint across nine domains — appetite and clearance, submission, quote and terms, bind and issue, service and claims on the distribution side; risk and exposure, bordereaux, authority and referrals, money on the capacity side.

Layer 02 · The record

One schema, mapped in once from each participant and mapped out once to each counterparty, replacing the point-to-point web where every new relationship is a build. It holds the relational, changing shape of a risk rather than one value per field, with lineage back to source on everything it carries.

Layer 03 · Workflows

Event-driven automation running across the systems already in place. Triggers on submissions, documents, schedules and changes in the record; parsing, classification and extraction with confidence scoring on every field; validation that gates every write; exceptions cleared by a person, with the fix promoted into a rule so the same break doesn't return.

Layer 04 · Agents & Applications

Where the work actually happens, built on the three layers underneath. Applications are full products running inside Noor and sharing its connections and its record — Noor's own and third parties' — installed from a marketplace rather than implemented over months. Agents are the same thing without a screen: they read the record, call the endpoints and run the workflows under the same permissions a person would hold, taking on the chasing, checking and reconciling that currently fills somebody's day. Both are bound by the layer beneath them, so neither can do anything the business hasn't authorised.


What runs on it

The same seven stages, with the layer underneath them.

01 · Digital transacting for brokers

Appetite, clearance, quote, bind and service endpoints on top of systems that never had them — with binding limits, territory and referral triggers checked on every call, so brokers integrate once instead of emailing a submission.

02 · Submission intake and triage

Documents parsed, classified by line of business and extracted field by field with confidence scoring, straight into the record — with only the uncertain fields reaching a human.

03 · A risk record that stays current

Attributes updated on events rather than on a cycle, enriched from third-party data, with claims and service activity written back to the risk they belong to — and able to trigger an action when something moves: do not renew, request an MTA, flag for review.

04 · Bordereaux, as two live ledgers

Decompose the monthly file into a Transaction Ledger for money and a Risk Attributes Ledger for exposure, both read from the same record. Premium recognition stops waiting on an exposure correction.

05 · Exposure capacity can actually use

Carriers, reinsurers and capital providers get a view of what they are carrying rather than a file describing it: exposure at the grain each one specifies, updated as the risk changes rather than reassembled monthly, and traceable back to the record it came from. Delta delivery, not full-file resends.

06 · Money that moves on time

Financial events booked as they happen, append-only and attributable, reconciled against what was billed and what actually settled — and pushed to the platforms that move the cash.

Every one of these runs on the same connections, the same record and the same engine. Which is the point: the first one is a project, and everything after it is configuration.


Why now

Something is about to start asking these systems questions.

Buyers, brokers and carriers are all putting AI in front of their processes. Within a couple of years a meaningful share of what arrives at an MGA won't be a person filling in a form — it will be software asking, on someone's behalf, whether there is appetite for a risk, what the terms are, and whether it can bind. Retail insurance is already being distributed this way. Specialty is harder, and not for the reason people assume.

Software can't transact with a business it can't address. It can't be trusted by a business with no record of what's true. And in a delegated market it can't go near a bind without an authority model to say what it may and may not do. That's three problems, and they are the three layers Noor already is.

Appetite, in machine terms

An MGA can publish what it will and won't write as a structured, callable answer, so an AI knows in a second whether to bring it the risk — instead of finding out three emails later.

Transaction, not conversation

Clearance, submission, quote, bind and endorsement as MCP endpoints on the systems already in place. Every call grounded in the record, validated before it writes, and answered the same way whether the caller is an AI, a broker's system or a person.

Authority, enforced

Binding limits, referral triggers, licensing and delegated authority checked on every request, with an immutable trail of what was asked, what was answered, and on whose authority it was agreed.

The businesses that capture this traffic will be the ones that are addressable when it arrives. Being addressable is not a project to start once it does.


Why Noor

Built so the market can adopt it without agreeing on anything first.

  1. 01 · Nothing has to change

    Noor is a layer, not a replacement. The PAS stays. The processes stay. The team keeps working the way it works. Every previous attempt at this asked a fragmented market to move as one, or asked one participant to carry the cost of a problem another one feels. Noor integrates around the stack instead, so the value arrives before anyone has to be convinced of anything.

  2. 02 · The second thing is faster than the first

    Because everything runs on the same connections and the same record, the work compounds inside your business and across the network. The integration built for one carrier is available for the next. The extraction rule written once applies everywhere the field appears. What starts as an integration project ends as a configuration screen.

  3. 03 · Neutral by design

    Noor doesn't compete with the systems it connects, doesn't own the customer relationship and doesn't take a position in the risk. It is infrastructure, and it is only useful if every participant can trust it equally — so each one keeps its own view of the business, its own vocabulary and its own commercial edge.


The standard

The standard should be an output, not an entry requirement.

Common templates, ingestion utilities and shared data models have all been tried, in reporting and in every other corner of this market. Each governs the format the data arrives in. None of them govern how it's captured, maintained or corrected, which is where the work actually is — so the work stays. Noor takes the opposite position: hold one record that can carry the complexity, and any standard the market settles on becomes a view the layer renders on demand. Core Data Record. A carrier's own template. An MCP schema an agent expects. Whatever replaces all three. You don't join a standard through Noor. You produce them.

For the participant

Keep your systems and your vocabulary. Meet every counterparty's format without maintaining a mapping for each one.

For the counterparty

Take the grain you need, at the cadence you need, without asking anyone upstream to re-platform to give it to you.


Ready to see Noor in action?

Twenty minutes, your systems on the whiteboard, and an honest answer on whether this fits.