Training course · AI Agent Development · Legacy Modernisation

AI Agent Development for Legacy Modernisation

Build AI agents professionally: architectures, tool use, agent SDKs, LangGraph, memory, evals, guardrails and deployment.

Duration
2 days (14 hrs)
Format
Online or on-site
Language
German or English
Participants
open dates or exclusive for your team
Certificate
Certificate of attendance

Overview

“AI Agent Development for Legacy Modernisation” is designed for roles such as Developers maintaining COBOL, PL/I or RPG systems, Java and .NET teams with ageing monoliths and outdated frameworks and Software and enterprise architects in modernisation programmes and shows step by step how to use AI Agent Development productively, safely and transparently in Legacy Modernisation – from the fundamentals to advanced workflows.

AI agents go beyond conventional chatbots: they break tasks down, call tools and APIs, evaluate the results and keep iterating until a goal is met. In this two-day course you learn to build such systems so that they work reliably, transparently and safely in day-to-day business. You compare architectural patterns ranging from simple workflows through the agent loop to multi-agent systems, and you implement them hands-on with the Claude Agent SDK, the OpenAI Agents SDK and LangGraph.

The focus is not on a quick demo but on the path to production. You design tool interfaces and context deliberately, connect tools via the Model Context Protocol, plan short-term and long-term memory and build evaluations that surface regressions early. On top of that come guardrails against prompt injection and harmful actions, human-in-the-loop approvals, tracing, cost control and deployment. By the end you can judge when an agent is worthwhile, when a deterministic workflow is enough and which framework fits your project.

Many business-critical systems still run on mainframe COBOL, ageing Java EE applications or .NET Framework solutions whose original developers moved on long ago. AI coding assistants and agents make these codebases far more approachable: they explain programme logic, reconstruct business rules, generate characterisation tests and support stepwise migration to modern languages, frameworks and cloud platforms. Successful modernisation is still engineering work, though, because automated translation without understanding simply produces new legacy code. In this course you learn to apply AI along proven patterns such as strangler fig, branch by abstraction and anti-corruption layers, lock down existing behaviour with tests before every change and base modernisation decisions on a solid analysis of what you already have.

The course is delivered live – either online in a virtual classroom or on-site in our training rooms in Munich, Salzburg, Berlin, Augsburg, Hanover, Bremen, Stuttgart and Kaiserslautern. We can also run it exclusively for your team.

Who is this course for?

  • Developers maintaining COBOL, PL/I or RPG systems
  • Java and .NET teams with ageing monoliths and outdated frameworks
  • Software and enterprise architects in modernisation programmes
  • IT managers and project leads responsible for migrations
  • Application modernisation and cloud migration consultants

Typical challenges we address

  • Business rules are buried undocumented in code and the original developers’ knowledge is gone
  • Missing automated tests make every change risky
  • Outdated frameworks, unpatched libraries and proprietary runtimes
  • Big-bang migrations fail on scope, data migration and parallel operation

What you will be able to do

  • You choose the right pattern for a task, from a fixed workflow to an autonomous agent.
  • You design tools and context so that agents work reliably and cost-effectively.
  • You implement agents with the Claude Agent SDK, the OpenAI Agents SDK or LangGraph.
  • You set up evaluations and tracing to measure quality and detect regressions.
  • Use AI agents to understand legacy systems, document business rules and make dependencies visible.
  • Protect existing behaviour with generated characterisation tests before migrating or refactoring code.
  • Plan and steer incremental modernisation using proven patterns instead of risky big-bang cutovers.

Course outline

Day 1

Module 1 Agent architectures and design patterns
  • Workflows versus agents: prompt chaining, routing, parallelisation, orchestrator-worker
  • The agent loop: plan, act, observe, reflect
  • Multi-agent systems, subagents and handoffs between agents
  • Criteria for choosing model, framework and level of autonomy
Module 2 Tool use and context engineering
  • Function calling and structured outputs with JSON Schema
  • Designing good tools: granularity, naming, error messages, idempotency
  • Connecting and reusing tools via MCP servers
  • Managing the context window: summarisation, compaction, retrieval and prompt caching
Module 3 Frameworks in practice
  • Claude Agent SDK: agent loop, tools, subagents, permissions and hooks
  • OpenAI Agents SDK: agents, handoffs, guardrails and tracing
  • LangGraph: stateful graphs, checkpointing and human-in-the-loop
  • Side-by-side comparison on the same task and criteria for choosing a framework
Module 4 Memory, evals and observability
  • Short-term and long-term memory, storage strategies and forgetting
  • Evaluation datasets, LLM-as-a-judge and task-based success metrics
  • Tracing tool calls, token usage and latency
  • Regression testing when prompts, models or tools change

Day 2

Module 5 Guardrails, security and deployment
  • Protecting against prompt injection, data exfiltration and over-privileged tools
  • Input and output checks, approval steps and sandboxing
  • Deploying as a service, asynchronous execution, retries and cost limits
  • Governance, audit logging and requirements under GDPR and the EU AI Act
Module 6 Practice lab: Legacy Modernisation
  • Extracting business rules
  • Mapping dependencies
  • Generating characterisation tests
  • Framework upgrades
  • Translating and refactoring code
  • Designing a strangler fig architecture
Module 7 Data protection, quality & transfer to daily work
  • Legacy systems often process customer, financial or insurance data: you learn to sanitise code and test data before using AI, clarify licensing and copyright questions around translated code and address the requirements of the GDPR, DORA and sector regulators for changes to critical systems.
  • Ground rules for safe AI use in your team
  • Checking results: facts, sources, tests and the four-eyes principle
  • Personal action plan for the next 30 days
  • Open questions, outlook and resources for further learning

Practical scenarios from your field

Extracting business rules

Analyse COBOL paragraphs, copybooks or Java service classes, document the embedded calculation and decision logic as readable business rules and validate them together with domain experts.

Mapping dependencies

Reconstruct call graphs, JCL job chains, database access and interfaces from source code and visualise them as diagrams to find sensible seams for stepwise replacement.

Generating characterisation tests

Capture current behaviour with golden master and approval tests, derive anonymised test data from production patterns and build a safety net before any migration begins.

Framework upgrades

Plan migrations from Java EE to Jakarta EE or Spring Boot and from .NET Framework to modern .NET, identify API breaks and apply consistent changes across many files.

Translating and refactoring code

Turn COBOL modules into idiomatic Java or C# rather than translating line by line, map data types such as packed decimals correctly and verify the result against the original with comparison tests.

Designing a strangler fig architecture

Design facades, routing layers and anti-corruption layers through which new services take over legacy functions step by step, including data synchronisation and fallback options. Each stage remains independently deployable.

Possible hands-on projects

Copybook to schema converter

A tool that reads COBOL copybooks and generates JSON schemas, Java records or SQL tables, correctly mapping decimal and REDEFINES fields.

Legacy knowledge portal

A small web app that makes analysed programmes, business rules and dependencies searchable and answers questions about the legacy system with source references.

Comparison test harness

A test harness that sends identical inputs to the old and new system, compares outputs field by field and logs discrepancies clearly.

Prerequisites

Solid programming skills in Python or TypeScript, experience with REST APIs and some hands-on experience with LLM APIs or AI assistants. You have programming experience in at least one language of the legacy or target platform, such as COBOL, Java or C#, plus a basic grasp of software architecture.

Technical requirements

A laptop with Python or Node.js, Git and a code editor, plus API access to at least one LLM provider such as Anthropic or OpenAI. Credentials for the exercises can be provided if needed.

Data protection & compliance

Legacy systems often process customer, financial or insurance data: you learn to sanitise code and test data before using AI, clarify licensing and copyright questions around translated code and address the requirements of the GDPR, DORA and sector regulators for changes to critical systems.

Suggested dates

All dates are suggestions – your preferred date is always possible.

  • Thu, 26 November 2026 – Fri, 27 November 2026 Online Request
  • Tue, 8 December 2026 – Wed, 9 December 2026 Berlin Request
  • Tue, 15 December 2026 – Wed, 16 December 2026 Online Request
  • Tue, 12 January 2027 – Wed, 13 January 2027 Stuttgart Request
  • Mon, 25 January 2027 – Tue, 26 January 2027 Online Request
  • Thu, 11 February 2027 – Fri, 12 February 2027 Hanover Request

Frequently asked questions

Which agent framework should we use?

It depends on your model strategy, the level of control you need and your team’s skills. Vendor SDKs offer a lean start with tightly integrated features, while LangGraph gives you finer control over state and flow. In the course you solve the same task with several frameworks and derive clear selection criteria for your own project.

How do you stop an agent from taking wrong or dangerous actions?

With several layers of protection: narrowly scoped tools with minimal permissions, input and output validation, approval steps for critical actions, sandboxing and continuous evaluation. Prompt injection via external content is treated as a realistic threat. You practise these measures directly on your own agent.

Does every automation need an agent?

No. Many tasks are handled more reliably, cheaply and transparently by a fixed workflow of a few LLM calls. Agents pay off when the solution path cannot be planned in advance and the system must react flexibly to intermediate results. We train exactly this trade-off using practical examples.

Can AI automatically translate a COBOL system into Java?

AI translates individual modules far faster than manual work, but a pure one-to-one translation often yields code that is hard to maintain. A better approach is to understand and test the business rules first and design the target architecture deliberately. In the course you combine AI-assisted translation with comparison tests and refactoring.

How do we handle very large codebases?

Large legacy systems do not fit entirely into a model’s context. What works is systematic inventory, per-module summaries, dependency graphs and targeted retrieval of relevant code. Agents then work module by module with a clearly defined brief. You learn to build this structure and turn the results into input for architecture decisions.

How much does “AI Agent Development for Legacy Modernisation” cost?

The course costs a flat €949 plus VAT per participant – whether online or on-site. This includes 2 training day(s) with a live trainer, course materials, a lab environment and a certificate of attendance.

Can I suggest my own date?

Yes. Our suggested dates are just a guide. In the request form you can always describe your preferred date or period in free text – we check availability and confirm within one working day.

Is the course online or on-site?

Both are possible: live online in a virtual classroom or on-site in Munich, Salzburg, Berlin, Augsburg, Hanover, Bremen, Stuttgart or Kaiserslautern. We can also run the course exclusively for your team.

More courses for Legacy Modernisation

More AI Agent Development courses

All 30 AI Agent Development courses

AI Agent Development · MunichAI Agent Development · SalzburgAI Agent Development · BerlinAI Agent Development · AugsburgAI Agent Development · HanoverAI Agent Development · BremenAI Agent Development · StuttgartAI Agent Development · KaiserslauternAI Agent Development · Online