Munich Datageeks January Edition

We are thrilled to announce our next Meetup on January 27th at Unit8.
Format:

• 2 talks (each ca. 40 min incl. discussion)
• Time for networking + food + drinks before, in between, and after the presentations
• Talks are held in English
• We will be taking photos and/or film footage at the event. These will be used to share news about our meetups and to publicize upcoming events.

The lineup:
First talk:
Mirella Haldimann & Frederic Becker - How to build Artificial Research Intelligence (ARI) Platform from zero to production ready in 6 months

Abstract:
The presentation will cover the development cycle of a greenfield agentic platform build to an actual product application. We will cover success factors, challenges, architecture design and implementation, benchmarking, etc. while also shedding light on the importance of business understanding for the involved engineers.
Bio:
Mirella Haldimann leads as Data & Consulting Lead thru complex projects for Unit8 clients. She has a Phd in Business & Engineering from Linköping University and ETH Zurich. She is also an accomplished Senior Data Scientist with a deep expertise of complex data architectures.

Frederic Becker is a Data AI Engineer at Unit8 with a Master in Cognitive Science from the University of Tübingen and focusing on RAG and domain specific GenAI Applications.

Second talk:
Mark Sandal - The Design Document: A Dull Beginning for a Drama-Free Ending

Abstract:
Writing documentation rarely tops anyone’s to-do list. It can feel slow and unnecessary—right up until poor communication, unclear goals or mismatched expectations turn into weeks, months and in some cases years of rework, frustration and endless meetings only to then culminate in an entire project getting tossed because it never addressed the real need.
This talk is about how one deceptively simple tool, the design document, can save your ML project from that fate. We’ll cover how it turns fuzzy ideas into concrete plans, exposes problems before they snowball, and keeps everyone aligned as things evolve. Along the way, you’ll see how starting with a few “boring” pages can be the difference between a smooth delivery and a costly post-mortem

Bio:
I’m a Senior Applied Scientist at Wolt, where I work in our Logistics team to make sure you get your food as fast and as affordably as possible.
In my (still relatively short) career, I’ve checked off pretty much the whole corporate data science starter pack — churn models, ranking, time-series forecasting, image classification, and more.
And if from all of this I had to pick one thing I wish the younger version of me knew in advance, it would be this: a ML/AI-related project’s success or failure is dependent first and foremost on proper communication between everyone involved – which is where the design doc comes in 🙂