Designing software architectures for AI and ML systems

Dates and Booking

Take your software architecture skills to the next level and learn how to develop robust, customizable architectures for ML and AI systems.

In our practice-oriented training course “Software Architecture for AI and ML Systems (SWARC4AI)”, you will learn how to successfully integrate ML components into modern software architectures. You will learn what is important when designing long-lasting, scalable and secure architectures for AI-supported systems - from data acquisition and processing to the use of LLMs, operation and further development.

The training is aimed at software architects and developers who have already gained initial experience with machine learning and now want to learn how to embed ML systems sustainably in software architectures.

Our SWARC4AI training is not a frontal event - you will work intensively on practical case studies. In small teams, you will formulate ML problems, prepare data, develop features, select models and train or adapt them. You will exchange ideas with other participants, discuss possible solutions and benefit from different perspectives. Our trainers support you with impulses, questions and assistance - for independent and sustainable learning.

Agenda

• Integration of machine learning systems into existing software architectures
• Designing architectures for data acquisition, processing, persistence and utilization
• Dealing with architectural requirements such as scalability, maintainability and ethical responsibility
• Overview of different ML and LLM models and their impact on architecture
• Recognizing and dealing with problems such as model drift or data bias
• Integration of LLMs into architecture concepts
• Legal, ethical and security-related aspects when using AI systems

Your Benefits

Audience

This training is just right for you if you:

• have some experience with machine learning (e.g. simple model training in Python),
• have a basic understanding of software architecture, DevOps and APIs,
• are confident in Python and typical libraries such as scikit-learn, pandas or PyTorch,
• want to develop architectures that sustainably support ML systems.

Your Trainers