Skip to content

Claims, policies and recoveries, with agents that start from your code and ask first

Policies, claims and payments live in different systems, and every payment has someone accountable. Reelease reads the code of those systems (COBOL and CICS, PL/SQL, Oracle Forms and Reports), and the agent cross-checks the data and prepares the decision. The person in charge approves it.

A use case from your sector, with sample data. Sign in with your email. We email you the link to open on your computer.

Claims paid sooner, policyholders who renew

60%
of policyholders not fully satisfied with their claim point to how fast it was settled.1
30%
of them have already switched insurer.1
250,946
compensation requests reached Spain’s Consorcio after the 2024 DANA floods.2

No payment goes out without a person

Agents on your policy and claims systems, with the approvals you define (human-in-the-loop) and what DORA asks of your ICT providers.

  • Every payment, approved

    • Payments wait for the claims lead
    • By default, whoever starts a task can’t approve it
    • Amounts and coverage, in the run record
  • No lock-in

    • Replace your policy system and you switch the connection
    • You choose the AI model
    • Exportable run records and knowledge
  • In the EU or in your network

    • Service in Madrid, in the EU
    • Or in your cloud or on your premises
    • Policyholder data queried, not copied

DORA, article by article

How Reelease helps with ICT third-party risk. The assessment is yours; the controls and documentation are ours.

Art. 28
ICT third-party risk: you get the documentation to assess us and for your register of information, and you choose where Reelease runs.
Art. 29
Concentration risk: you choose the AI model and where Reelease runs, including your own infrastructure.
Art. 30
Contracts and exit: exportable run records, knowledge and documentation, and agents that stay the same when you replace a system.
See the run record of a task

Four tasks, under control

Illustrative cases of what an agent would do with Reelease in this sector. The agent only uses what your team has published.

Home claim

  • Policies
  • claims
  • payments
Situation

A home claim comes in and the policy and coverage must be validated before paying.

Step by step
  1. Agent: Reads the notice and checks policy and coverage

  2. Agent: Calculates the payout and its rationale

  3. Person: The head of claims approves the payment

  4. Agent: Releases the payment and tells the policyholder

Control: The calculation and its rationale stay in the run record.

What you gain

The payment is ready as soon as the last document arrives, and the policyholder doesn’t have to call.

You pay sooner without skipping a single control, and the policyholder has one more reason to renew.

On your policies, mainframe included

Policies and claims on the mainframe, COBOL or in-house systems, payment APIs, databases through the Client Bridge and Microsoft 365, Google Workspace or Jira.

  • IBM Z mainframe

    • COBOL
    • CICS
    • JCL
    • PL/I
    • IMS
    • VSAM
    • Db2 for z/OS
    • IBM MQ
    • z/OS Connect

    Acts through Core transaction and Web API

  • Java and the JVM

    • Spring
    • Quarkus and JAX-RS

    Acts through Web API

  • Web, scripts and APIs

    • OpenAPI and Swagger
    • NestJS
    • Python
    • Ruby on Rails
    • GraphQL
    • React, Vue and Svelte

    Acts through Web API

  • Business tools

    • Microsoft 365
    • Gmail
    • Jira and Confluence
    • Azure DevOps
    • LinkedIn
    • Slack
    • GitHub

    With Each operator's work account and Organization app

  • See how they connectThe full technology catalog and how the agent acts on each one.

Trademarks and logos belong to their respective owners. They are shown to identify integrations and do not imply affiliation or endorsement.

Try insurance operations live

Step into a policy and claims system, with sample data and its agents ready to work: the main one and one per function.

  • Ask about policies, coverage and claims.
  • Choose a home claim: the agent checks the cover, calculates the settlement and gives you the letter to the policyholder.
  • See the payment wait for your approval, and review the run record.
Try it

A use case from your sector, with sample data. Sign in with your email. We email you the link to open on your computer.

Frequently asked questions

Can the agent pay a claim?

It can prepare the payment. By default, the person in charge approves it, never the one who launched the task.

What about policyholder data?

You decide what the agent can look up at each step, with each person’s work account. What it reads goes only to the AI model you choose, and no provider trains on it; the details are on the Security page.

Does it work if policies live on the mainframe?

Yes. Reelease analyzes COBOL, CICS, PL/I and JCL as they are, and your team decides which tasks the agent can run, with nothing to rewrite.

See it on your own system

Pick a task, we analyze it in your system and you try it in a real 3-month pilot, with hands-on help from Reelease.

You choose the cookies

We use technical storage to remember your choice and, only if you accept, analytics cookies (Google Analytics) to understand how the site is used. You can change your choice at any time. Learn more in our cookie policy.