Home/Mindbreeze/Use cases/Claims and customer service
Process library · Insurance and financial servicesClaims and customer service: one answer with sources, across your systems
A customer asks about a claim. The answer sits in the policy terms, the claim file, earlier e-mails and the CRM. Instead of reconstructing it from four screens, the handler gets one answer with links to each source, and decides.
Discuss this use case See an example
In short
In this proposed scenario, claims handlers and customer service agents at an insurer or financial services company ask a question in their own words and get one answer assembled from approved policy information, the claim or case file and earlier correspondence. Every statement links to its source, and each person sees only what they are allowed to see. The handler checks the answer and takes the decision. SoftAdvice sets this up with Mindbreeze on top of the systems the company already uses.
The knowledge is there, just not in one place
A customer calls about their claim. The handler opens four systems while the customer waits.
Claims handlers and service agents answer questions all day: is this covered, what is the status, what did we already tell this customer? The facts exist, but they are spread over the policy administration, the claims system, document storage, mailboxes and the CRM.
So every answer is a small research job. Experienced colleagues are interrupted with the same questions, new colleagues need months to know where to look, and incoming mail is read and routed by hand before anyone can start.
- Policy terms and product conditions
- Claim or case file
- Earlier e-mails and letters
- Customer data in the CRM
From searching four screens to checking one answer
The handler keeps the decision. What changes is the time spent finding the facts behind it.
Today
- Searching system by system
- Unsure which policy version applies
- Experts interrupted with the same questions
- Incoming mail sorted by hand
In this scenario
- One question, one answer with sources
- Approved documents ranked first
- Shared knowledge, fewer interruptions
- Incoming items classified and routed
From customer question to a verified answer
Mindbreeze does the finding and assembling. A person checks and decides.
Question arrives
By mail, letter or phone. Incoming items can be classified and routed to the right team before anyone reads them.
Ask in context
The handler asks in their own words, from the case they are working on.
Answer with sources
Mindbreeze assembles an answer from the policy terms, claim file and correspondence the handler may see, with a link to each source.
Verify and decide
The handler opens the sources where needed, decides and answers the customer. Nothing is sent automatically.
One question, the sources behind it
An illustrative question from a home insurance claim. The answer names the policy version that applied on the date of the damage, the status of the file and what was already promised to the customer, each with its source.
- Only documents the handler is allowed to open
- Approved policy information before drafts
- Every statement traceable to a source
Illustrative example
The person decides, the sources stay visible
An answer is a starting point for the handler, never a decision on a claim. Because every statement links to its source, a colleague or an auditor can see later what the answer was based on.
Mindbreeze applies the permissions of each source system: what a handler cannot open in the source, they do not see in the answer. If access in a source is too broad today, that is fixed in the source; we check it before the start.
- Answers support, people decide
- Permissions of each source respected
- Sources linked in every answer
- Access checked before the start
What we measure on a first use case
We agree the baseline with your team before the start, so the result is measured on your own questions.
Time to answer
How long a handler needs for an agreed set of typical questions, before and after.
Answer usefulness
Handlers rate answers against the agreed questions.
Source traceability
Whether every answer points to the right, current source.
Permission tests
Test users confirm that nobody sees what they may not open.
What this scenario needs, and what it does not do
The answer is only as good as the sources. Before the start we agree which documents are the approved version, which systems are connected and which team owns the first use case. Connectors, configuration and licensing are confirmed per system.
This is a proposed scenario. It is based on patterns in customer cases that Mindbreeze has published for insurance and customer service, described here without names or figures. It is not a SoftAdvice implementation reference.
- Agreed approved sources
- Connectors confirmed per system
- No automatic claim decisions
- One team, one use case first
Questions about this use case
Does the AI decide on claims?
No. The answer helps the handler find the facts faster. The handler checks the sources and takes the decision, as today.
Can a handler see documents they normally cannot open?
No. Mindbreeze applies the permissions of each source system. If access in a source is too broad, that is corrected in the source; we check it before the start.
Do we need to move our documents?
No. The documents stay in the systems where they are today. Mindbreeze connects to them; which systems can be connected is confirmed per system before the start.
Is this only for insurers?
No. Any service team that answers questions from policies, contracts, case files and correspondence works this way, from financial services to public service desks.
How do we start?
With a conversation about one use case: the team, the typical questions and the systems involved. From there we propose a scoped first step.
Which questions take your team the longest today?
Tell us about the team, the questions and the systems involved. We come back with a proposal for a scoped first step.
