Bootcamp content
NLP basics and key technologies
We start with an in-depth introduction to the most important NLP techniques such as tokenization and vector representation. Based on these concepts, you will learn how probabilistic models serve as a basis for text processing and analysis. You will then apply these concepts in practical, real-life exercises and build the skills to work with modern language models.
LLMs and Agentic AI
The second part of the bootcamp focuses on transformer-based language models such as GPT. You will learn the basics of prompt engineering and retrieval-augmented generation (RAG) for context optimization and use your new skills to develop advanced LLM applications. One highlight is the development of so-called Agentic AI systems. They are LLM agents that can autonomously structure tasks, operate tools and make decisions. You will learn to use these agentic systems to build productive, adaptive and interactive AI solutions for real-life tasks in business, research and administration. You will also learn about advanced concepts of agent orchestration and multi-agent workflows: How can multiple agents coordinate your work? How can you design robust agentic flows for complex business processes? From sequential task management to parallel agent systems, you will build a deep understanding of how to orchestrate intelligent workflows.
Model context protocol (MCP)
You will familiarize yourself with the model context protocol (MCP), an open standard that facilitates the efficient and consistent access of LLMs to external tools, data sources and APIs. MCP creates a uniform interface through which LLMs are reliably supplied with relevant context information – irrespective of its source.
Security, fairness and red teaming
It’s not enough for modern NLP systems to perform as desired, they also have to be safe, robust and trustworthy. That is why we will take a deep dive into topics such as bias recognition, fairness, transparency and MLOps. In practical exercises, you will learn to apply so-called red teaming strategies to test LLMs for weaknesses, undesired output and safety risks and to establish effective guardrails for your models.
In just eight days, this bootcamp will take you from the theoretical foundations to practical application, with a focus on future topics including AI, MCP and responsible AI development. After the bootcamp, you will be able to help shape the future of language AI.
