Automating an insurance broker's work

The needs questionnaire is filled in by hand for every policy, and its data is retyped into requests and recommendations. During an inspection, it is hard to find the complete set of documents.

The system prepares a questionnaire pre-filled with client data, and once completed it creates the needs analysis document and saves it with the policy. The broker's recommendation refers to the same analysis.

The broker's power of attorney, company details, property schedules and existing policies are collected over a series of emails. Some documents arrive weeks later.

The new client receives a form and a list of documents to upload. The system pulls company details from registers, sends reminders about gaps and creates a client record with the end dates of current policies.

Clients ask about instalment dates or claim status, or request a certificate of insurance for a tender. The broker answers emails instead of analysing risks.

A chatbot or email assistant answers based on CRM data: instalment dates, policy cover, claim status. It sends certificates automatically and passes questions that need advice to the broker.

The same client data is entered into several insurers' forms, each with a different layout. A request to five insurers takes several hours.

From the client record, the system prepares requests in the formats required by each insurer, attaches schedules and claims history and sends them to the selected insurers. Replies come back to one case.

An issued policy sometimes differs from the accepted quote: a different sum insured, a missing clause, an error in vehicle data. This only comes to light when there is a claim.

When the policy is received, AI compares it with the accepted quote and client data: sums insured, limits, clauses, period and subject of insurance. Differences go to the broker with a ready correction request to the insurer.

Insurers change their policy wordings, and the broker needs to know what has changed in exclusions and limits. Comparing two versions of several dozen pages takes hours.

AI compares the new version of the policy wording with the previous one and points out changes in definitions, exclusions, limits and the insured's obligations, with section numbers. The broker reviews the summary and sees which clients are affected.

Policy end dates are in the CRM, but preparing a renewal starts when the broker remembers the client. With a large portfolio, some renewals are done in a rush.

90, 60 and 30 days before the policy ends, the system creates tasks, asks the client for up-to-date data and launches quote requests. Every week, the broker sees a list of renewals with the stage of each one.

The client forgets an instalment, and the insurer sends a demand. The broker finds out about the problem when cover is already at risk.

The system knows the instalment schedule of every policy, reminds the client before the due date and checks payment information from the insurer. Unpaid instalments go to the broker before the matter becomes serious.

The client buys new cars or opens a warehouse and reports it by email, sometimes weeks later. Additional cover and additions to the fleet policy are handled by hand.

The client reports a change through a form or email, AI identifies the vehicle or location and prepares the notification to the insurer. The system tracks confirmation and updates the schedule on the policy.

The client describes the loss by email, the broker asks for documents and forwards the claim. Status is checked by phone with the loss adjuster.

The client reports a claim through a form with a list of documents for that type of loss. The system passes the complete claim to the insurer, tracks deadlines, including the 30 days for payment from notification, and keeps the client informed of progress.

Commission statements from insurers arrive in different formats. Checking that commission was calculated on every paid instalment is done only occasionally.

Commission statements are imported automatically and compared with policies and paid instalments in the CRM. Missing or understated commissions go onto a list for follow-up with the insurer.

The client's loss ratio is calculated by hand before renewal, based on emails from insurers. There are no arguments for premium negotiations.

The system collects claims and payout data for each client and policy, calculates the loss ratio and prepares a report for the client. The same report goes into the quote request at renewal.