We design, develop and licence underwriting software for insurers, MGAs and brokers — the quoting, referral, binding, document and reporting machinery behind a book. Whatever the class, and whoever carries the risk.
Underwriting is being rebuilt around machine assessment. Most of the market is waiting to see how it settles. We are building it.
Our leadership team brings more than forty years in the insurance industry — broker side, insurer side and MGA side, underwriting risks and running delegated authorities.
Most insurance software is built by technologists who have had the business explained to them. It shows, and always in the same places: the question that does not quite fit the risk, the referral nobody can evidence six months later, the schedule that says something different from what was agreed, and the bordereau that gets rebuilt in a spreadsheet every month because the system will not produce what the carrier actually asked for.
We build from the other direction. We have sat in the underwriter’s chair, held the pen on a binder, argued with a carrier over a reporting format, and explained a declined risk to a broker with a client waiting on the phone. We have also worked across a good many of the systems already in this market, which is a fast way to learn what a platform must do and what it must never make people put up with.
A complete underwriting platform — risk capture, rating, referral, binding, documents, adjustments, renewals and reporting. One system, one record, one version of the terms. The engine is deliberately indifferent to what you write: property, liability, commercial combined, marine, specialty. The class is configuration.
Question sets that adapt to the risk in front of them, so a submission arrives complete rather than half-answered.
Rules-driven pricing and terms applied consistently every time, with the straightforward business quoted without a queue.
A clear line between what binds automatically and what an underwriter must see, with the reason recorded rather than implied.
Schedules, statements of fact and certificates produced from the decision itself, so paperwork cannot drift from the terms.
Adjustments, cancellations and renewals as first-class operations, each re-rated and re-evidenced properly.
Every decision, override and change attributable to a person and a moment, because eventually somebody will ask.
Machine assessment now sits inside mainstream UK underwriting rather than alongside it. Our platform scores property and location risk from external data at the point of submission, and returns that score as part of the risk — not as a separate report somebody has to go and read.
The score is applied through the same rule set as everything else. It can price a risk, load it, or send it to an underwriter. It does not bind and it does not decline on its own.
An underwriter sees the score, the factors behind it and the data it was drawn from, and can agree, override or set it aside — with the reason recorded against the risk.
An AI-influenced outcome has to be explainable to a carrier, a regulator and a policyholder months later, and shown to have been made by a person. The system is built so that record exists by default.
Most platforms treat every product as a development project, which is why launches take quarters. Ours separates the engine from the product configured on top of it.
The question set, rating basis, what binds and what refers, and the documents it must produce.
1–3 daysConfigured to your appetite, with quotable cases early so you react to something real.
1–2 weeksYour underwriters run genuine risks through it, including the awkward ones.
About a weekLive with a few brokers first, watching real submissions, before wider release.
Days to adjustThe platform is rarely the problem — the book is. Nobody wants to rekey several thousand live policies mid-term while still writing new business, so the decision gets deferred year after year.
Terms carry across exactly as written. We do not re-rate or re-decide historic business on the way in. Current rules apply at the next renewal or adjustment.
Existing policies are read by machine from the documents you already hold and reconciled policy by policy against their own source. Anything that does not agree is flagged for a person. You finish with a report you can put in front of your carrier.
We licence our software to insurance businesses under a written agreement. We do not sell it, and we do not transfer ownership of it.
A right to use the software for an agreed purpose, for an agreed term, by an agreed number of users.
All intellectual property in the software remains exclusively vested in Apex Digital Risks Ltd at all times.
Data entered by a licensee remains the licensee’s property, used only to provide the software and related services.
You do. Data entered by a licensee remains that licensee’s property, and we use it only to provide the software and related services to you. It is never used to benefit another client.
It moves across without being rekeyed and without being re-rated. Policies are read from the documents you already hold and reconciled one by one against their source, with anything that does not agree flagged for a person. Terms carry across exactly as written; current rules apply at the next renewal or mid-term change.
No. A score can price a risk, load it, or send it to an underwriter. It cannot bind and it cannot decline on its own. Where data is missing or unavailable the risk refers to a person rather than defaulting to an assumption.
Yes, from the record rather than from memory. The score, the factors behind it, the data it drew on, the rule that acted on it and any override with its reason are all held against the risk. Rules and scoring behaviour are versioned, so you can state what the platform was doing on the date a policy was written.
For the build, yes — because the engine, the portals and the document production already exist, and a product is configuration on top. What genuinely takes longer is carrier sign-off on rates and wordings, legal review, and your own team’s capacity to test with real cases. Start those on day one, in parallel, or you will wait months for a product that was ready in weeks.
No. Apex Digital Risks is a software company. We are not authorised or regulated by the Financial Conduct Authority and we do not arrange, advise on, underwrite or sell insurance. Regulated activity carried on using our software is the responsibility of the authorised firm conducting it.
Quote a risk in your browser and watch it decide, refer, or fall outside appetite — premium, terms, endorsements, audit trail and schedule all moving together. Then tell us what you actually write.