Speed Meets Architectural Competence
Without an architectural framework, however, that’s mostly fast failure. Technical debt and governance gaps arise unchecked, and can later only be fixed with great effort. In this seminar, you’ll learn to make architecture decisions with AI support and prepare architecture knowledge so that it actually works in agentic contexts.
Agenda
How LLMs work, prompt and context engineering, the agentic loop
Extracting architecture knowledge from existing systems
Using AI support to gain architecture information from code, interfaces and dependencies
Preparing architecture knowledge for agents
Architecture rules, context files and retrieval mechanisms such as knowledge graphs, LLM wikis, skills, RAG and MCP
Keeping agents within architectural boundaries
Preventive and detective controls: architecture tests, linting, review gates
Making architecture decisions with AI
Translating quality goals into drivers, generating hypotheses and design variants, and deciding with confidence
Governance, legal and security aspects of AI use
Copyright, data protection, EU AI Act, traceability, sandboxing
The architecture discipline in transition
Preserving judgment, recognizing biases, rethinking roles and collaboration
Your Benefit
You learn to shape architecture work effectively in an agent-driven development environment. You make architecture decisions with AI support, data-driven, and as testable hypotheses. You can prepare architecture knowledge so agents can use it at the right moment. You can design and implement preventive and detective controls to deliberately enforce architecture decisions. You extract architecture information from existing code and turn it into living documentation. You classify governance, legal questions, and security risks of AI use, without slowing down your development speed. Even with high automation, you retain technical and substantive responsibility for the overall system as an architect.
Target Audience
This training is aimed at software developers and software architects whose company uses, or plans to use, generative AI and agents in development. You should bring practical experience in software development or architecture, as well as a basic understanding of architecture concepts, modern development processes, and build pipelines. Helpful, but not required: first practical experience with a coding agent and basic knowledge of architecture decisions, ADRs, or CI/CD.
iSAQB® Accredited
In our seminar, we implement the learning goals of the iSAQB® curriculum AGENTA in a modern, hands-on and interactive way. On your way to becoming a Certified Professional for Software Architecture (CPSA) Advanced, you receive 20 methodological and 10 technical Credit Points.
Our Extras for You
All exercise results and theoretical content shown are made available to participants electronically afterwards
Official Certificate of Participation
In this seminar, participants are awarded 20 methodological and 10 technical Credit Points for the iSAQB® CPSA-Advanced Level with a perpetually valid iSAQB® certificate.