LLM Deep dive: Part 1 @ Magenta Telekom

Dear Data Enthusiasts,

The recent break throughs of LLMs and generative AI have hugely influenced the way we think about machine learning and it’s potential applications. To address the potentials and challenges that come with such paradigm shift, we have decided to organize a focus series about in the upcoming months. In the upcoming Part 1 @ Magenta Telekom, you will learn how LLMs are already used today and how you can make the most of it!

We are super excited to present you with two incredible speakers who can provide an expert perspective on the matter, going beyond the hype and buzzwords.

Zoom for online attendance: https://ceu-edu.zoom.us/j/96857211595?pwd=WkVPWkJuQ29BVUhBeVgzNWdMZFFJZz09

Don’t just engineer prompts, engineer workflows! Klaudius Kalcher, Entrepreneur and Angel Investor

Despite the increasing prevalence of LLMs in our daily lives and most users' increasing awareness of their capabilities and weaknesses, many workflows are still built on a simple single-query approach. In the meantime, the tooling to build systems that cater more to the strengths of the underlying models has improved immensely, ranging from libraries to no-code platforms, making it easy to build systems using even the most sophisticated methods. In this talk, we will provide a high level overview of some typical state of the art designs like Retrieval Augmented Generation (RAG) and self-correction loops, and explore practical application using a real-world example.

Klaudius Kalcher is an entrepreneur and angel investor focusing on AI ventures. He received a PhD in Medical Physics in 2014, working on machine learning analysis tools. In 2017, he co-founded the synthetic data company MOSTLY AI, where he held the role of Chief Data Scientist until 2022. Since then, he has been active as advisor and angel investor.

Let’s build our own ChatGPT in 30 minutes! Zoltan Toth, Freelance MLOps Consultant

What if you wanted to build your custom and private LLM Chatbot locally? In this live demo, you will see how you can build an AI chatbot from scratch in no time using local LLMs, Huggingface, and ChainLit. We’ll grab an LLM Model like Meta’s Llama2, convert it into a chat agent, and build a simple UI around it..live on stage. The coding environment and the code will be made available after the talk.

Zoltan is a freelance Data Architect and MLOps Solutions Architect. He has worked on topics ranging from Data Engineering through MLOps to Scalable Machine Learning Systems, delivering projects for several Fortune 500 clients globally. He also teaches big data computing at CEU and is a board member of VDSG.

🎤🎤 Open Mic We are going to open up the stage after the talks for community announcements. If you’d like to announce something, open this slide deck, make sure you are signed in with a google account, and click “View Only” -> “Request Edit Access”. Explain in the text box what you want to announce, and we’ll give you edit access to the slide deck. 🎤🎤 Also, please note that during the event, photos might be made and later posted on VDSG’s or Magenta’s social media page. Please notify us if you do not agree. Attention attendees with food allergies. Please be aware that the food and drinks provided may contain or come into contact with common allergens, such as dairy, eggs, wheat, soybeans, tree nuts, peanuts, fish, shellfish, or wheat.