Why AI and LLMs shouldn’t actually work and why they do anyway
BAYOOTEC CTO David Ondracek shows in the webinar on 20.05.26 how Large Language Models, Retrieval Augmented Generation, GraphRAG and autonomous agents really work. Step by step, without buzzword bingo, with clear explanations and concrete practical examples.
Teile diesen Beitrag:
Teilen Sie diesen Beitrag:
From tokens to RAGs and autonomous agents: How large language models really work
Large Language Models do not understand meaning. They know no truth. They statistically predict the next word. Nevertheless, they write texts, program complete apps, answer questions and solve complex tasks. How does that fit together?
In this webinar, David Ondracek, CTO at BAYOOTEC, shows how modern AI systems work under the hood. Step by step, he goes through the central building blocks: tokens, embeddings, transformers, retrieval augmented generation, GraphRAG and finally autonomous agents.
Comprehensible examples are used to illustrate how pure statistics can be turned into powerful tools, where the limits lie and where their use in everyday working life is really worthwhile.
The webinar is aimed at anyone who not only wants to use AI, but also wants to understand it. Without buzzword bingo, but with clear explanations and concrete practical examples.
What participants take away from the webinar
- A clear understanding of how LLMs actually work, without any prior knowledge of math
- Orientation in the jungle of terms: tokens, embeddings, RAG, GraphRAG, agents, neatly categorized
- The ability to distinguish AI hype from real added value
- Concrete starting points where AI can be usefully applied in everyday working life
- A look at regulatory frameworks such as the EU AI Act, which helps with project decisions
About the speaker David Ondracek
As CTO of BAYOOTEC GmbH, David Ondracek is responsible for the technological direction of the company and in particular drives the use of AI in software and business processes. His current focus is on the development and implementation of sustainable AI strategies.
He has been working intensively with large language models, retrieval augmented generation and agent-based systems for many years and has accompanied numerous AI projects, from the initial feasibility study to productive operation. As a certified AI compliance officer (TÜV), David combines technical know-how with a clear view of regulatory frameworks such as the EU AI Act, risk classes and compliance requirements.
How we engineer secure AI solutions for companies
Artificial intelligence only unfolds its value when it is properly embedded from a technical and regulatory perspective and tailored to the specific use case. At BAYOOTEC, we combine in-depth engineering expertise with a clear view of the EU AI Act, risk classes and compliance requirements, from the initial feasibility study through RAG and GraphRAG to agent-based systems in productive operation.
Request now
Would you like to implement an AI project, request David as a speaker or have general questions about individual software engineering? We are happy to support you and look forward to hearing from you.
