AI Agents For Architects

Architectures For Agent-Based AI Systems

Format: Online Training
Duration: 2 day

Direct enquiries

Architectures for agentic AI systems

Agentic AI systems plan tasks, execute steps and coordinate tools. This training combines strong foundations with practical takeaways and shows how to architect and classify agentic AI systems. It focuses on the core building blocks of agentic systems and the key questions that should be addressed early in both design and operations. Upon completion of the training, participants will be able to assess agentic AI initiatives in a structured way and confidently justify architectural decisions, from the initial assessment to readiness for production environments. This training course is aimed at anyone who wants to efficiently integrate AI solutions into existing IT landscapes while focusing on aspects such as scalability, security and maintainability.

What can you expect?

The AI Agents for Architects training provides hands-on architectural knowledge for developing a structured understanding of agentic AI systems and positioning them appropriately in real-world initiatives. Participants learn to view agents as a combination of core building blocks and to understand the role of context, memory, and state in enabling reliable behavior. The training also emphasizes interoperability, including standards such as MCP, along with quality and observability, and the security and scalability requirements of production IT landscapes.

Prerequisites for Participants
E

Solid understanding of software architecture and the design of software systems, APIs, and integration interfaces

E

Grundverständnis von GenAI und LLM Nutzung, zum Beispiel Kontextfenster, Tokens sowie RAG als Konzept

Your Added Value
E

You understand the agentic paradigm and can clearly distinguish agentic AI systems from classic GenAI scenarios.

E

You learn how agents are architecturally designed and how their core building blocks can be embedded in a resilient overall system.

E

You will acquire practical decision-making logic for context engineering as well as memory and state management.

E

You will learn how tool integration and interoperability are structured in agent architectures, including in the context of MCP and multi-agent systems.

E

You will learn criteria and procedures for evaluating the quality, robustness and security of agentic systems.

E

You will learn how to prepare agentic systems for productive environments, with a focus on observability, scaling and secure operation.

Training content in detail

From Model to Agent System

➤ Classification of what characterizes agent-based AI systems and why this changes architectural decisions.
➤ Overview of the key building blocks that interact within agent-based systems.

Context, memory and state under control

➤ The role of context, memory, and state in ensuring reliable behavior, and common pitfalls to watch out for.
➤ Approaches for managing context so that systems remain stable under real-world conditions.

Tool connection and collaboration in agent landscapes

➤The role of context, memory and state in reliable behaviour, and where typical pitfalls lie.
➤Approaches to handling context so that systems remain stable under real-world operating conditions.

Making quality visible

➤How agents use tools and why interoperability should be built into the design from an early stage.
➤ MCP and multi-agent systems as guiding concepts for standardized collaboration and extensibility.

Ready for Production

➤ How quality can be meaningfully conceptualized and evaluated in non-deterministic systems.
➤ Observability as a foundation for making system behavior traceable in operation.

Open dates and registration

-100€

up to 6 weeks before the start of the training.

Are you interested in this training?

Please call us at +49 621 595702 41, send an email to Academy@itech-progress.com, or submit a written inquiry using our contact form:

Our Trainers

Your benefits at a glance

E

Hands-on content: The training focuses on real-world examples and practical exercises.

E

Modern concepts: You gain a structured overview of key approaches to agentic AI systems and their architectural principles.

E

Efficient integration: A clear roadmap for integrating and operating agentic AI solutions in existing system landscapes.

Technical Requirements

No special setup is required for the training environment. Exercises are conducted using cloud- and web-based tools such as Miro, Draw.io, and platforms like Jupyter or HuggingFace. A stable and sufficiently fast internet connection is required.

Live-Online-Training

E

Includes training materials and exercise resources

E

Visual collaboration using tools such as whiteboarding

E

High level of interactivity through tailored exercises and breakout rooms

E

Optimal trainer support, even in the breakout rooms

Back to Extended Training
Interested in iSAQB training?

These trainings might also be of interest:

Software Architecture for AI Systems (SWARC4AI)
Foundation for future-proof AI systems – SWARC4AI

AI4SWARC – AI as a Tool in the Architecture Process
AI4SWARC is a newly developed training on using AI as a tool in the architecture process.

Participant Feedback from the Training

“The training helped me understand the logic behind systems that independently break down tasks, coordinate tools, and maintain consistent results across multiple steps. I found the guidance on context, quality, and operations particularly valuable, as these are exactly the areas where real project questions arise.”

Participant from Cologne

Would you like customized in-house training?

Do you have an entire team you would like to train, but none of our courses meet your requirements? No problem! We would be happy to create a customized in-house training in collaboration with you, whether for beginners or advanced participants. We look forward to your inquiry!

Inquire Now+49 621 595702 41