Skip to main content

Agentic coding workshop for insurance companies

Insurers run policies that last decades, on systems that carry every tariff generation ever sold. Our workshop shows your IT and actuarial teams how AI coding agents help with exactly this kind of long-lived, rule-heavy code, safely and under supervision.

ISO 9001 & ISO/IEC 27001Certified quality and information securitySince 2014, made in GermanyDiese Seite auf Deutsch

Why insurance code needs a specific approach

A life or health policy written twenty years ago still has to be administered correctly today, which means old tariff rules, special clauses and product versions live side by side in the policy administration system. Much of this knowledge exists only in code and in the heads of a few experienced developers. AI coding agents such as Claude Code, OpenAI Codex and Cursor can read this code, summarize business rules in plain language and write tests that lock in current behavior. That makes them valuable for documentation and modernization, provided the team knows how to verify what the agent claims.

neooutput, the AI education unit of jocapps GmbH, tailors the workshop to insurance scenarios: extracting tariff rules into readable specifications, migrating actuarial scripts from spreadsheets or VBA into tested Python, building claims intake services and reviewing interfaces to brokers and comparison platforms. Sessions run remote or on-site, in English or German. We also address the regulatory side, from DORA and supervisory IT requirements to the special protection of health data.

What to watch out for

  • Health and life data is a special category under Art. 9 GDPR. Make sure no real policyholder records, medical information or claim files ever end up in an agent's context, not even as debugging examples.
  • Agents can describe tariff rules convincingly and still be wrong. Every extracted business rule needs confirmation by someone from the product or actuarial side before it is treated as a specification.
  • Treat the AI provider as an ICT third-party service. Under DORA insurers need the same assessment, register entry and exit planning as for other critical IT services.
  • Calculation code that feeds Solvency II or financial reporting needs documented validation. Agent-assisted changes do not replace the model validation and governance processes you already have.
  • Plan for mixed tool landscapes. Policy systems, mainframe components and modern services often coexist, and the agent setup has to respect which repositories are in scope.

What your insurance IT gains

Business rules in plain language

Agents turn nested tariff logic into readable descriptions that business departments can confirm. Knowledge that sat in a single developer's head becomes documented and reviewable.

Safer modernization

Your team learns to generate characterization tests before any refactoring. Changes to premium or reserve calculations then fail loudly instead of slipping into production.

Actuarial code under control

Actuaries and developers practise moving calculation logic from spreadsheets into versioned, tested code. The result is easier to audit and to reproduce for reporting.

Governed AI beyond IT

For underwriting or claims teams that want AI support too, we show how a platform like kalAIDoCreate adds policies, logging and custom APIs. That keeps sensitive use cases out of private chat accounts.

How the workshop is organized

  1. Scoping call

    We discuss your core systems, languages, approved tools and the line of business you want to start with.

  2. Sample code preparation

    We prepare sanitized or synthetic code that mirrors your tariff, claims or interface logic.

  3. Hands-on sessions

    Developers and actuaries work with the agent on realistic tasks, always with verification and review built in.

  4. Adoption roadmap

    You receive usage rules, review checklists and a proposal for a first pilot in one product line.

Matching products

neooutput

Agentic coding workshops and AI services. Hands-on agentic coding workshops for teams on Claude Code, OpenAI Codex and Cursor, remote or on-site, in English or German. Plus AI Code Engineers on demand who deploy, secure and scale AI-built apps.

Explore neooutput workshops

kalAIDoCreate

Governed AI platform for enterprises and the public sector. An all-in-one AI platform for enterprises and public institutions with compliance features, governance and custom APIs.

Explore kalAIDoCreate

Frequently asked questions

Can AI agents work with COBOL or other legacy languages?

They can read and explain much legacy code and help write tests or migration drafts. Quality varies by language and code style, so we practise verification techniques rather than assuming the output is correct.

Should actuaries attend a coding workshop?

Often yes. Many calculation models live in scripts that actuaries maintain themselves, and they benefit from learning how to let an agent add tests, documentation and version control to their work.

How do we start without risking production systems?

We recommend a pilot on documentation and test generation first. Those tasks deliver value without the agent changing productive logic, and they build trust in the review process.

Is the workshop available in German?

Yes. All sessions are available in English or German, remote or at your location.

Let's talk about your project.

15 minutes with an AI Code Engineer, free and without obligation. Or call our team around the clock.