WEBINARS
AI-Bots for Academic Institutions
AI-Bots are particularly well suited for supporting research.
Cologne University of Applied Sciences (TH Köln), together with North Rhine-Westphalian Library Service Centre (hbz), has been using the ASK bot for several months – first in an intensive testing phase and, since December 2025, in live operation.
In this webinar, we will share the first experiences of a university library working with an AI-Chatbot. Using practical examples from everyday library work, we will show which special requirements must be considered in an academic context.
We will also demonstrate different use cases:
- how to interact with large collections of text files and PDFs using the Document-Bot
- what kind of information the Service-Bot can provide
- how the Recommendation-Bot supports research and academic work
Decision Maker Webinar
on AI-Chatbots For Libraries
AI-Chatbots are taking over the world traditionally reserved for Search. Even google puts an AI – Response before providing the search results.
ASK has developed the first AI – Chatbot for public libraries, providing recommendations on library titles and answering service questions since June 2024. After nearly 1 year in service it’s time to provide insight into our experiences on user satisfaction, costs and efficiency.
Previous webinars of our series have focused on explaining AI. This webinar will focus on how to use AI chatbots in a library. Our goal is to provide decision makers with all the information they need to assess if an AI chatbot can be of any use for their library. We will address the shortcomings of AI as well as its strengths to provide a balanced (albeit slightly biased, of course) picture.
We will address the following topics, among others:
- How do users interact with an AI chatbot? How many are there and how satisfied are they?
- What are the target groups of AI-Bots and what is the social impact?
- Are there alternatives to ChatGPT? How does Mistral perform? What about the local hosting of an LLM?
- How does the library staff respond to AI? Is there any fear of AI taking jobs?
- What about concerns on data-protection and privacy?
- What does it take to implement a bot? What does the library need to do?
- How much does it cost?
AI-Chatbot for Libraries and Institutions
Prompting And Conversations. Part I.
This webinar will explain the concepts and the terminology defining AI. We will also provide practical examples and a live demo on how these concepts impact our everyday work with our AI Chatbot for libraries.
When the internet arrived, we had to understand a whole new set of words and concepts. “Email”, “WLAN”, “Bluetooth”, etc. have become household terms. It is easily forgotten that they had to be learned. Until we know about “LLM”, “Tokens” or “RAG” we will have no real understanding of how to think about and how to use AI. We will lift the veil.
After this webinar you will be able to answer these questions:
- What is RAG and will it make searching for books and other media more reliable?
- What is a token and what does it cost?
- Is “open source” really open source?
- Why do we need a Vector Database for our metadata? What is a vector anyway?
- Are there metrics for a good AI – conversation?
- Why is an AI hallucinating? Does it matter?
AI-Chatbot in Public Libraries
First Practical Experiences
A webinar on the first experiences we made running Europe’s first AI Chatbot for the Association of Public Libraries in Berlin (VÖBB).
On 28th May 2024 we have launched the first AI-Chatbot in Europe for public libraries. Its goal is to complement traditional search, allowing patrons to chat about ebooks, films, music, etc.
In this webinar we explain why the Berlin libraries made their approx. 4 Mio media available using an AI chatbot. We provide provide real data and practical experience on:
- User response to chatting about library collections
- Working with diverse, inconsistent and incomplete data
- Ownership of data and what to do about it?
- Prompting your chatbot for library use
- Pricing: on the cost for infrastructure, customization and tokens
- A word on ethics and data protection
Speakers:
- Moritz Mutter, Head of Digital, VÖBB Service Center, Association of Public Libraries in Berlin
- Tilman Scheel, CEO, AboutSomethinK
You will find our Chatbot on the website of Berlin libraries voebb.de.
Info-Bot For Libraries and Institutions
This webinar will show you what AI service bots can do – and how. We will also show how to use the service bot for internal knowledge management (which is essentially the same).
We will address the following topics:
- How to chat with a service bot about information provided by the library (live).
- What sources could be used? PDFs, Wikis, Websites, etc.?
- How can I best ensure the quality of my data?
- How to prompt service questions?
- Can I use the service bot for knowledge management inside my institution?
- Can’t ChatGPT do it? Or Google? Or someone else?
- Again a word on hallucinations.
- What does it cost and why?
- Will Service Bots “steal” jobs?
The webinar includes the live prompting in the backend of our service-bot, as well as real experiences of some of our clients.
AI-Chatbot for Libraries And Institutions
The Technology
This webinar will focus on the technological aspects of AI-Chatbots.
Nevertheless, the webinar will also be understandable for those who only occasionally deal with AI and technology.
As always, we will provide practical examples and a look under the hood of our chatbot.
We will mainly focus on two technical aspects of an AI chatbot:
- Issues rooted in the nature of AI vs issues that can be solved using prompts, function calls, better data, etc.
- Trade-Offs: Speed vs. advanced LLMs, hosting local models vs. API, etc.
AI-Chatbot for Libraries and Institutions
Prompting And Conversations. Part II.
In this webinar we will focus on the eloquence of AI-Chatbots: What can chatbots do that “classic” keyword search cannot? What is a helpful conversation? How can you measure the quality of an answer? We will chat with our bots on questions that will test its limits. We will provide live insight into our prompting backend.
We will show you how we measure quality and what we do against hallucinations. Throughout the presentation we will be transparent with the Chatbots’ shortcomings. And above all we will allow a peep into the (not so distant) future.
Topics are:
- Good prompting: How do I explain to the bot what it is supposed to say?
- Metrics: How do I even know what a good chat is?
- Hallucination: What if the bot is talking nonsense?
- Talking about images or films: the multimodal bot.
- Can the bot also speak Farsi, Japanese and Spanish? At the same time?
This is a second part in our webinar series on the concepts and terminology behind AI-Chatbots.