In this edition of our ScaDS.AI Meetup, we provide insights into two highly topical fields of AI application:

1. We look at sovereign generative AI solutions for the public sector, which are used in particularly security-critical and regulated environments such as public administration.
2. We present an AI research agent that was developed specifically for scientists to make research in millions of publications more efficient and well-founded.

Look forward to two inspiring talks with live insights into technology, application, and benefits – and then take the opportunity to exchange ideas with our guests, the ScaDS.AI Dresden/Leipzig team, and other interested parties from the region.

Talk 1: Sovereign GenAI for public administration – with the ]init[ AI Platform

Speaker: Tom Schlansky, Key Portfolio Manager for AI at ]init[ AG

How can generative AI be used safely, scalably, and in an application-oriented manner in highly sensitive environments such as public administration?

Tom Schlansky provides an exciting insight into the ]init[ AI Platform – an open-source-based GenAI platform developed specifically for sovereign use cases in public administration. The platform has a modular structure (LLMs, VLMs, agents), is based on microservices, and can be operated in BSI-certified high-security environments. In addition to a compact overview of the architecture and security aspects, a specific use case will also be presented. Time for questions and discussion is expressly planned.

Talk 2: AI Research Agents: scienceOS in research

Speaker: Henri Max Deda, co-founder of scienceOS

Henri Max Deda introduces an AI tool that was designed for scientific research and knowledge acquisition.

ScienceOS is the AI research agent for researchers with high standards and little time. With access to over 225 million scientific publications, scienceOS provides well-founded answers with precise references. Features such as PDF uploads, a personal library for uploading and sharing up to 4,000 documents, and other intelligent tools make working with scientific literature easier. Complex algorithms ensure that reliable sources are prioritized. In this talk, Henri Deda demonstrates the features of scienceOS using four typical challenges in everyday research:

• from quickly familiarizing yourself with new topics,
• to the targeted comparison of publications,
• the identification of particularly relevant sources and fundamental works,
• to outlining and reviewing your own texts.

See how scienceOS’s AI Research Agents can change everyday scientific life with real-world application examples.

Agenda

• 4:45 pm – Doors Open
• 5:00 pm – Welcome & Intro
  Presentation of ScaDS.AI, the Living Lab & the Agenda
• 5:10 pm – Talk 1:
  Sovereign GenAI for public administration – with the ]init[ AI Platform
  (Tom Schlansky, ]init[ AG)
• 5:40 pm – Talk 2:
  AI Research Agents: scienceOS in research
  (Henri Max Deda, scienceOS)
• 6:05 pm – OpenLab & Networking
  Exchange with the speakers, insight into our demo setups, and open discussions in a relaxed atmosphere

Model-Context… what?
In this hands-on edition of the monthly ScaDS.AI Meetup, participants will explore a promising innovation in the world of large language models (LLMs) and their interaction with the outside world, which is already being rapidly and widely adopted: the Model Context Protocol (MCP).

What is MCP? Why was it introduced, and how is it changing the way AI systems manage tools, memory, and state? In our AI Insights Talk, we’ll unpack the core ideas behind the protocol, explain its purpose, and explore the value it brings to developers and AI users alike.

Afterwards, you’ll get the chance to explore and use MCP yourself – whether you’re a curious beginner or an experienced developer.

Agenda

AI Insights: What’s the Model Context Protocol?

This session will cover:

• • What is the Model Context Protocol – and why now?

• • How MCP connects memory, tools, functions, and external data.

• • What makes it different from earlier approaches like tool calling or prompt engineering?

• • Why MCP is quickly becoming a core part of modern AI ecosystems?

• • What practical value does MCP offer to developers, users, and companies and their processes?

You’ll leave with a clear understanding of the architecture behind MCP – and why it possibly matters for the future of AI-based applications.

Interactive Workshop: Use the Model Context Protocol

Time to roll up your sleeves: In the workshop, we’ll get to know and work with MCP in two tracks:

• • Beginner-Friendly Track (Low/No Code): Use Langflow based on prepared examples to visually and practically see MCP in action – no code required.

• • Developer Track (Code-Based): Dive into a real codebase and implement MCP-based workflows yourself using a prepared repository, templates & examples.

Whether you want to experiment or implement, we’ve got you covered.

What to bring:
No setup required – we’ll provide ready-to-use laptops on-site. But feel free to bring your own device if you want to work on your personal setup.

OpenLab and Networking

After the workshop, join us in our OpenLab area to meet fellow attendees, share your ideas, and ask questions. Whether you’re new to the topic or already experimenting with MCP-based systems, this is your chance to connect with the community and researchers behind the scenes.

Join us for an evening of context, code, and collaboration – and get hands-on with the Model Context Protocol!

How do we use artificial intelligence in practice today? What steps are involved in implementing your own AI idea? And how can companies or individuals get started? Join us and the KI-Hub Sachsen-Thüringen at the next ScaDS.AI Meetup, a networking event for anyone interested in AI or looking to develop their own ideas.

This time, we will focus on three practical perspectives: of start-ups, an established company, and support services from research institutions and their transfer into practice. Find out what the journey from idea to functioning AI product can look like, even without programming skills. Using a local company as an example, we show how medium-sized companies integrate AI into the day-to-day business, products and services. You can also talk to organisations that support the transfer between research and application with their services.

Whether you are looking for initial points of contact with the topic of AI or are already pursuing your own projects, this Meetup offers space for insights, exchange and new impulses.

Agenda

Welcome & introduction

Oliver Welz, ScaDS.AI Team Service & Transfer / Living Lab

A brief presentation of ScaDS.AI, its research, services and points of contact for companies and interested parties.

Introduction of the KI-Hub Sachsen-Thüringen

Martin Folz, KI-Hub Sachsen-Thüringen

What does the KI-Hub offer? Presentation of central support services, transfer formats, and success stories from cooperating with companies, especially SMEs.

Talk 1: Immocheckonline: a real-life AI use case with almost no budget or programming skills

Jan-Philipp Schubert from Immocheckonline.de

Jan will demonstrate how he developed a fully automated solution for verifying property documents in just three months using low-code tools, AI models, and a clear use case. He implemented this solution with virtually no budget and little prior knowledge. In his presentation, Jan openly shares what worked well and what he would do differently today. This is an honest experience report for anyone who wants to set up an AI product themselves.
It is an honest experience report including learnings, setbacks and concrete tools.

Talk 2: VoluLink: Digital Impact Projects for Social Engagement with Future-Oriented Skills

Moritz Hall from VoluLink

VoluLink connects teams of IT students with experienced corporate mentors to work on IT projects for the common good. Co-founder Moritz Hall will share insights into the organisation’s mission and previous projects, and will also seek advice from the community on upcoming non-profit AI projects. Their vision: Technology for social change — together, we can harness the power of digitalisation to shape the world of tomorrow.

Talk 3: ISL Integrated Skills Deutschland GmbH (ISL): Future-proof route planning: Focus on efficiency, safety and satisfaction

Hubertus Kraus, ISL Integrated Skills Deutschland GmbH’s Managing Director, on site
Peter Pauckner (Technical Director, Integrated Skills) via MS Teams

How can costs be reduced through the targeted use of resources while ensuring service quality? How can occupational safety and driver satisfaction be increased within the company? – Quick and safe training of new drivers

Live demo: How we optimise your tours, even in difficult road conditions.

OpenLab: Networking & Deep Dive

Speakers will be available for discussion about AI, experiences, tools, business use cases and cooperation ideas. Take this opportunity to network in a relxed atmosphere over drinks.

We look forward to seeing you!

Accelerating Protein Design with Computational AI Methods

How can the latest AI methods be combined with biotechnological research to accelerate medical innovations? The ScaDS.AI startup AI-Driven Therapeutics (AIdT) shows how a start-up idea from research became reality – exclusively presents its first product: the “Co-Scientist”, an AI-supported chatbot to accelerate protein and drug development.

This event is aimed at anyone interested in the connection between AI and biotechnology – especially researchers, founders, students and industry representatives. You can look forward to exciting insights into the start-up journey, the technological implementation and the application possibilities of the co-scientist – as well as the opportunity to work with the technology yourself in the workshop.

Agenda

We look forward to your participation!

The ScaDS.AI Meetup as part of the Data Week 2025 takes place on June 11, 2025 from 5:30 pm to 7:00 pm at the City Hall in Leipzig.

Find more information about the event on the official Data Week Website.

Registration

As this Meetup is part of the Data Week Leipzig, registration is required.

Registration for the Meetup is also possible when purchasing a Data Week ticket for June 11.

Agenda

In the first part of the evening, three guests from research and industry will explain how AI can be successfully put into practice. From 7 pm, there will be a demo session that invites you to try out and discuss.

Introduction & Welcome

Oliver Welz (ScaDS.AI (Center for Scalable Data Analytics and Artificial Intelligence) Dresden/Leipzig)

Opening of the evening with a brief introduction to the agenda, introduction of the speakers and presentation of the ScaDS.AI AI Competence Center.

Live Fusion and Seamless Integration – DBpedia Enterprise in action

Dr.-Ing. Sebastian Hellmann (Institute for Applied Informatics)

In this presentation, you will learn how DBpedia Enterprise links heterogeneous data sources such as Wikipedia, Wikidata and Linked Open Data – in real time and with stable, searchable knowledge graphs. The technology not only enables effective data integration for companies, libraries and research institutions, but also presents a sustainable business model in the open data environment. Learn from concrete application examples how DBpedia can be used as a knowledge backbone and what role it plays in the “Web of Data”.

AI in practice – insights from everyday business life

Alex Hofmann (MaibornWolff GmbH)

How do we as a technology company manage to successfully implement AI projects? In which use cases does generative AI really bring an advantage? What AI blockades do we experience in our daily work and how do we solve them? The fact is: the AI potential is huge and affects all industries. The presentation will show interesting examples from sectors such as tourism, insurance and robotics. MaibornWolff, an established IT service provider with a focus on sustainable digitalization and AI, will report on its own practical experience.

Graph databases for AI applications – use cases with Neo4j

Dr. Christopher Rost (Neo4j, Inc.)

Graph databases are essential for many AI applications that focus on relationships between data. Neo4j, one of the leading providers in this field, will present how companies can use graph technology to analyze complex relationships, identify patterns and make informed decisions. The presentation will provide practical insights into fields of application such as fraud detection, recommendation systems and knowledge management.

Demo: Pamemo

Contact: Adrian Viehweger and Sebastian Driesnack.

“See one, do one, teach one” is the motto of Pamemo. The AI-supported learning app enables practical and interactive education on a smartphone from the comfort of your own home. It is currently being used in Leipzig and Africa to train doctors – primarily on the topics of resuscitation and antibiotic therapy.
Unlike simple wrappers around AI models, Pamemo combines various technologies to provide reliable and verifiable content – even in sensitive areas of application.
Supported by SAB validation funding, Pamemo is open to cooperation with universities, educational institutions and partners who train and educate specialists.

Demo: alphaspeech voicebot – Customizable voicebot framework for data-secure voice interaction

Contact: Dr.-Ing. Felix Gräßer [Product Owner Speech Technologies, alpha NT GmbH]

alphaspeech voicebot was developed for applications where data protection, flexibility and domain-specific adaptability are paramount. Voicebots can be optimally adapted to the respective speech domain, the application context and specific usage requirements. The modular framework is based on open source technologies and in-house developments in the areas of speech recognition (ASR), natural language understanding (NLU), dialog management and speech synthesis (TTS). By dispensing with the integration of third-party services, voicebot can be operated completely on-premises or on-device. This means that the highest data protection and data security requirements can be met.

Join us for the next interactive ScaDS.AI Meetup on May 20, 2025 and dive into the dynamic world of AI Agents and Agent Frameworks. Find out what they are and why everyone is talking about them right now. Learn what they can do – and where users should be careful.

In the course of the Meetup we will explore the latest developments, demystify the concept of agents, and show real-world applications and limitations. Participants will also have the chance to build their own agents in an interactive workshop – regardless of whether they have little or more experience with the topic.

Bring your own laptop if you want to try things out for yourself – we will guide you through the process step by step! The workshop is open to everyone. Programming knowledge can be helpful, but isn’t a necessity.

Program

AI Insights: Exploring AI Agents

This time, the focus lies on AI Agents and Agent Frameworks. The following questions will be discussed and answered:

• • What are AI agents, and why are they gaining so much attention?

• • How do they organize, distribute and orchestrate tasks?

• • What tasks can agents autonomously solve – and where are their limits?

• • What frameworks and tools make it easy to prototype your own AI agents?

• • How can users maintain control and security when agents interact with their environment?

We will have a look at typical use cases and demonstrate practical examples, giving participants a solid foundation to understand and work with AI agents. Feel free to bring your idea!

Interactive Workshop: Build Your First AI Agents

Join us for a hands-on session where we’ll help you build your first AI agents!

What to expect:

• • No setup needed – but feel free to bring your laptop to build and run your own agent!
    There will also be prepared laptops you can use.

• • No prior agent experience required – we will guide you through beginner-friendly tools.

• • Dive deeper if you want – we will explore more advanced agent frameworks for experienced users.

Workshop Structure:

5:45-6:00 pm Introduction, Setup & How to Start: Get started with beginner-friendly tools and/or set up everything you need to start coding your own AI agents.

6:00 – 6:35 pm Hands-On Practice: Create your own simple agents, automate tasks, and experiment with workflows.

6:35 – 6:45 pm Discussion: Let’s discuss the results: What did you achieve? What could be the next step? Will AI agents take over?

We will provide easy-to-follow guides and examples, so you can get started right away – and maybe even spark ideas for your own agent projects!

OpenLab & Networking

After the workshop, we invite you to join us for our OpenLab & Networking session. Meet fellow enthusiasts, researchers, and developers, share your experiences, and exchange ideas for new projects and collaborations. Participants who have already built or experimented with agents are encouraged to bring and present their experiences!

The LLM Programming and Understanding Meetup is a community of experts and enthusiasts in the field of AI programming. Our goal is to share knowledge, exchange new ideas, and collaborate on innovative solutions. Mr. Wilhelm Pfaffenroth will be the speaker for this evening and will provide you with in-depth insights into the topic of AI Agents. The agenda will focus on the following topics:

• • Introduction: A precise definition of the term AI Agents.

• • Potentials and limitations: An analytical look at the opportunities and challenges of this technology.

• • Use cases: Presentation of concrete application examples to illustrate the practical relevance.

• • Tech stack: A detailed presentation of the technological foundations and architectures.

• • Current development and future: Analysis of the latest trends and a forecast of future developments.

• • Discussion: An open discussion round in which you can put your questions to the speaker and the community.

How do you make the step from research to practice? How does an idea become a successful start-up? What support is available at Leipzig University for interested parties, researchers and students who want to found a spin-off? If you are interested in the topic of spin-offs and would like to gather valuable information, then this meetup is the right place for you! Get in touch with people who are already in the middle of the start-up process or can support you.

At the next ScaDS.AI Meetup on April 15, 2025 at 5:00 pm we will discuss central aspects of spin-offs, the topic of transfer at the University of Leipzig and show which offers and contacts are available through the start-up initiative SMILE.

A successful start-up will show the way the founders have gone and report on the possibilities and ways to implement a start-up idea using their own example.

Gain valuable insights from experts and practitioners, make new contacts and discuss with us how to transfer innovations successfully to the market.

Program

Knowledge and Technology Transfer at the University of Leipzig: Research with Impact – Bridge to Practice

Dr. Dirk Wilken, Head of the Office for Knowledge and Technology Transfer at Leipzig University

The Transfer Department presents itself, its work and central strategies to promote knowledge transfer. The focus will be on opportunities for cooperation between science and industry, securing and exploiting intellectual property rights, supporting spin-offs and best practices from transfer practice. You will gain a comprehensive insight into the topic of transfer at Leipzig University.

Start-up support: Support on the way to your own spin-off | Offers & Opportunities

Dr. Gunnar Kaßberg, SMILE Start-up Initiative of Leipzig University, International Start-ups

What advice and funding opportunities are available at the Leipzig University for people interested in starting a business? The start-up initiative SMILE of Leipzig University presents offers, programs, funding possibilities and success stories. What does it take to (successfully) found a company? What needs to be considered, where are the hurdles, where are the opportunities?

You will get practical insights and tips from the idea to the first planning of a spin-off to the next steps.

Startup Insights: From idea to successful start-up with AI-Driven Therapeutics

Fabian Liessmann, Co-Founder and Managing Director of AI-Driven Therapeutics GmbH

A successful start-up provides practical insights into its own founding process: How was the product and the business idea found, which challenges were overcome, which networks were used and which learnings can be passed on? Finally, it shows how the technology developed supports the biotech sector and what added value it offers companies.

OpenLab: Networking, exchange & open discussion

Exchange ideas with other participants, make valuable contacts and discuss your ideas and experiences in a relaxed atmosphere. Take the opportunity to ask specific questions to our speakers and enter into a discussion.

The next ScaDS.AI Meetup on March 11th will again be a hands-on edition, this time focusing on the topic of local Large Language Models (LLMs). We will answer and discuss questions such as: What does it take to run LLMs locally? How do open source models compare to the proprietary base models of the major AI vendors? What are the benefits and challenges of running your own LLM?

In addition to exploring the latest developments and discussing real-world use cases, this meeting will give you the opportunity to run your own local LLMs in an interactive workshop. The workshop is open to anyone interested. You can bring your own laptop to experiment with local LLMs. We will provide you with prepared hardware, including a dedicated GPU machine. This will allow you to explore larger models on site.

1. AI insights: Exploring local LLMs

With a focus on local LLMS, we address the following points:

• • What open-source models are available, and where can you find the right one for your use case?

• • How do you keep up-to-date with the latest developments in open-source LLMs?

• • What hardware and software do you need to run a model locally?

• • What are the advantages of local LLMs in terms of data privacy and customization compared to cloud-based solutions? Are there disadvantages?

We will show a live demo to illustrate how to get an LLM up and running with Ollama, along with helpful tools like open-webui for a productive and private local setup.

2. Guest Talk: AI-Driven Pipeline for Generating Draft Medical Reports in Myocardial Scintigraphy

Dr. Peter Werner will present his project on using local LLMs for the automated creation of draft medical reports in myocardial scintigraphy. Data privacy concerns often limit the use of cloud-based LLMs in sensitive fields like healthcare. Local LLMs offer a promising solution by ensuring that data remains secure and private. This session will highlight a real-world application, demonstrating how local LLMs are being experimentally used to streamline repetitive tasks, such as medical documentation.

3. Interactive Workshop: Local LLMs, but how?

Join us for a hands-on workshop where we will guide you step-by-step in setting up and experimenting with local LLMs.

What you can expect:

• • No prior knowledge is required! Feel free to bring your laptop for your own local setup.

• • Programming experience is helpful, but not essential. We’ll help you get started and introduce you to the basics.

• • Hands-on learning: work in a collaborative and enriching environment based on the Meetup community and ScaDS.AI staff

Workshop Structure:

• • Introduction & preparation (5:45pm – 6:00pm): We install Ollama and start a local model. After everything is set up, including open-webui, we can start playing around.

• • Hands-On Practice (6:00pm – 6:30pm): Test it out for yourself! Try out different models, experiment with your own prompts and explore how the models run on a more powerful GPU setup. Get a feel for what it takes to run these models and how powerful they are compared to those from the major vendors.We are happy to provide a step-by-step guide to follow along.

4. OpenLab and Networking

After the presentations and the workshop, we invite you to our Living Lab, where you can get to know our lab and discuss with other participants, AI researchers and developers.
Take this opportunity to exchange ideas, network with like-minded people and explore potential collaborations.
In addition, we encourage participants who have already experimented with local LLMs to share their experiences with the community!

On Feb. 18, we are introducing an interactive format of the Meetup where you can dive into the latest AI innovations and actively participate in an AI pair programming workshop. Like the Meetup itself, the format is aimed at AI enthusiasts and experienced developers as well as curious beginners and interested parties.

1 – AI Insights
In this part, we will explore what’s currently shaping the AI community, from groundbreaking reasoning models to chain-of-thought (COT) approaches. You will learn about the most exciting innovations, trends and use cases and how these models can lead to impactful applications.
We will also present a live demo to bring these concepts to life, illustrating how reasoning and COT models like O1 or R1 work in practice. Whether you’re interested in theory, implementation, or finding practical resources like GitHub repositories or Huggingface Modelcards or even spaces, this session provides a comprehensive overview of the latest in AI.

2 – Interactive Workshop: AI Pair Programming
Roll up your sleeves and join us for a hands-on workshop designed for participants of all skill levels. Whether you are a complete beginner or an experienced programmer, this workshop will inspire and challenge you!
What to expect:

• We provide laptops and examples, you just bring your curiosity and motivation.
• No programming experience required. We will guide you through the process and help you get started.
• Collaborative learning: We will pair participants in groups to foster knowledge sharing, matching experienced programmers with beginners to create an enriching learning environment.

Workshop Structure:

• Introduction: Learn what AI-pair-programming is all about and see a showcase with a demo case.
• Guided Example: Follow step-by-step instructions to understand how to program with AI, complete with frameworks, extensions, and tools.
• Hands-On Practice: Experiment with your own ideas or prepared examples. Whether you are trying out AI pair programming for the first time or want to deepen your knowledge, this is your opportunity to immerse yourself in the subject matter.

3 – Outlook & Feedback

We willl close the formal part of the Meetup by sharing details about the next event, including the topic and date, and gathering feedback from you. Let us know what you loved about this Meetup and what you’d like to see in the future!

4 – OpenLab and Networking

After the workshop, we invite you to our Living Lab for live AI demos and the opportunity to connect with like-minded individuals, AI experts, and potential collaborators. Enjoy a relaxed atmosphere to share ideas, discuss challenges, and expand your network.

Immerse yourself on 25.11.2024 in the fascinating world of Artificial Intelligence (AI) and discover how projects can be successfully realized from conception to practical implementation. At the upcoming ScaDS.AI Meetup, you will gain valuable insights into the outstanding regional and local research landscape that makes Leipzig, Dresden, Saxony and Central Germany a central hub for AI development. Learn how to collaborate with our center, the AI Factory, and partners like InfAI, and explore the diverse opportunities for developing and integrating AI solutions. Do you have specific ideas or visions? Discuss them with us!

Join us for exciting presentations, an inspiring AI success story from a regional company, and plenty of networking opportunities during our OpenLab session. Gain insights into current research topics, test demonstrators and be inspired by innovative impulses in the field of Artificial Intelligence.

1. Presentation by KI-Manufaktur
   KI-Manufaktur introduces itself as a partner for small and medium-sized companies and commissioned work.
2. Company presentation: TIQ Solutions
   Paper documents, manual effort and errors are still part of everyday working life, while the integration of different systems and interfaces is a complex matter. These obstacles not only make data collection and processing more difficult, but also affect the quality and speed of decision-making. In this presentation, you will learn how you can use artificial intelligence (AI) to optimize your paper-based processes and structure data efficiently. Using illustrative use cases and examples of intelligent document processing, we will explain the challenges and procedures for leveraging the added value of the technologies shown for companies.
   
3. AI Insights
   In this section, we show what is currently moving the AI community, what innovations and trends there are and how the latest models can be used for exciting use cases. In a small accompanying live demo, we also present one or two showcases in practice. What possibilities arise from this, what current models are available and what can you do with them? Which GitHub repositories implement exciting projects and how do you get it all up and running (yourself)? We would like to give you an insight into our daily work, but also venture a brief overview.
4. Review of the year and outlook
   We look back on the year, the past Meetups and would like to venture a look ahead to the new year 2025. We also welcome feedback on the format!
5. OpenLab and networking
   After the presentations, we invite you to join our Living Lab. Experience our AI demos live and interact with other attendees, AI experts and network contacts in a relaxed atmosphere.