Most talks about AI talk about the future. Our workshop at the HER•izon Summit showed what already works today. Live, with real data, in front of an audience.
At the HER•izon Women Summit 2025 in Bad Gastein, Soneo AI was not just the technology partner behind the website and booking platform. We also held a workshop on a topic that occupies us daily: AI agents.
The workshop was deliberately hands-on. No lecture about the future of AI, but a concrete look at what is already possible today. Plus a live example built specifically for this summit.
From chatbot to agent: the decisive difference
The most important aha moment in the workshop was the distinction between a simple chatbot and a true AI agent.
A chatbot answers. An agent acts. It plans steps, uses tools, accesses external sources, and makes decisions on the way to the goal. That is the leap from "answers my question" to "gets my task done".
“An AI agent is convincing when it doesn't talk about AI but simply does what is needed.”
Exactly this distinction decides in practice whether AI remains a nice gimmick or creates real value.
The basics: RAG and LLMs explained simply
So that AI agents don't remain abstract, we explained the building blocks behind them. Understandable, without assuming technical background.
We showed how a large language model (LLM) works in the first place. And what a RAG system is: retrieval augmented generation. In other words, how an AI accesses its own knowledge base instead of relying solely on what it was trained on.
“An LLM knows a lot. A RAG system knows what's in your documents. That difference decides the quality of the answers.”
This building block is the reason an agent can access current, company-specific information instead of merely reproducing general knowledge.
The practical example: an agent for the HER•izon website
The strongest part of the workshop was the live example. We showed live how we built an AI agent specifically for the HER•izon Summit.
The agent could do two things at once:
Query the current program. Which sessions take place when? Which speakers are where and when? The agent retrieved the website's program data in real time and answered directly. No static FAQ, but a system that knows the current data.
Retrieve the current weather. Anyone traveling to a multi-day summit in the Alps has a simple question: what will the weather be like? The agent accessed weather data and wove the answer directly into the context. A small feature that generated a lot of enthusiasm in the demo.
That was exactly the idea behind this example: showing that an agent is not abstract. It connects real data sources with natural language and delivers answers that are actually useful.
Why hands-on workshops work
The HER•izon Summit brings together women from business, media, politics, technology, and society. An audience that doesn't need an academic introduction to neural networks, but concrete answers to the question: what's in it for me?
That is exactly what our workshop format is made for. We explain the fundamentals just far enough to carry, and then show what is possible with a living example.
For Soneo AI, the collaboration was a good example of how AI technology isn't only relevant for large corporations or mechanical engineering. AI makes sense wherever real problems need solving. And sometimes that's a summit agent telling hundreds of participants when the next workshop starts and what the weather will be.
Want to show your team hands-on what AI agents can do today? We hold workshops that don't stop at theory but start from the concrete use case.
Book a free initial consultationFAQ
What is an AI agent workshop and what do you learn there?
An AI agent workshop teaches hands-on how AI agents work and how they differ from simple chatbots. Participants learn how agents plan steps, use tools, and access external data sources to solve tasks independently.
What is the difference between an AI chatbot and an AI agent?
A chatbot answers questions based on predefined information. An AI agent goes further: it acts independently, plans steps, uses external tools and data sources, and makes decisions on the way to the goal. An agent can, for example, retrieve program data and weather data at the same time.
What is retrieval augmented generation (RAG) in simple terms?
RAG is a technique in which an AI accesses its own knowledge base instead of relying solely on its training material. The system finds relevant information in documents or databases and uses it as context for precise answers.
How do you build an AI agent for a business website?
An AI agent for a website connects a large language model (LLM) with a RAG system and external data sources. This lets the agent access the website's content, retrieve real-time data, and answer user questions in natural language instead of providing only a static FAQ.
Who are Soneo AI's workshops suitable for?
Our workshops are aimed at teams and decision-makers who want a hands-on understanding of what is possible with AI today. We assume no technical background, explain the fundamentals clearly, and use concrete use cases to show the value AI agents create in everyday work.
