Data Science Retreat Demo Day #44

Data Science Retreat presents 5 Machine Learning prototypes and projects by Batch 44 participants.

Free to Attend!

Hey Berlin Data Folks!
We’re kicking off 2026 with our first data event of the year.
Join us for project demos, fresh ideas, great conversations, and of course—pizza & drinks on us 🍕🍻
Don’t miss it!

Agenda:
17:30 - Drinks and Networking
18:00 - Welcome & Introduction
Followed by Project Presentations

Project Ideas:
1. AI/ML - EV Battery Shape Classification
Project by Uma Maheswari Obbani

The project aims to automate the detection and classification of used electric vehicle (EV) batteries by shape — cylindrical, prismatic, or pouch — using AI/ML and computer vision techniques. The goal is to provide manufacturers and recyclers with real-time insights about battery type distribution, enabling efficient sorting, inventory management, and safety compliance.

2. Developing Scalable and Reliable Coffee Yield Forecasting Tools
Project by Dr. Juan Fernando Duenas Serrano

The aim of the project is short term coffee yield forecasting improves planning and value creation across the supply chain, from farmers to roasters. Current yield estimation methods are not scalable and require specialized expertise. This project explores computer vision as a scalable alternative using smartphone images. A YOLOv8 object detection model counts unripe Arabica coffee cherries on branches. The model can estimate yield at tree or farm level with optional user input. A Gradio or Streamlit web app will demonstrate, evaluate, and extend the model’s potential.

3. Classic Car Diagnostic Tool
Project by Lulezim Ukaj

This project aims to develop a specialized AI-powered diagnostic tool for classic cars that leverages community knowledge from German automotive forums. Unlike modern OBD-II–based diagnostic tools, which dominate the market, this solution targets pre-1996 vehicles where standardized electronic diagnostics do not exist.

4. Enhancing on-site audience audio
Project by Thede Witschel

This project develops an enterprise-grade AI platform that automates the extraction of ESG data, regulatory compliance checks, and peer benchmarking for companies. Utilizing NLP and machine learning, the system converts unstructured sustainability reports into standardized metrics, facilitating real-time compliance monitoring and competitive intelligence across various industries. Business Impact: Targets the rapidly growing ESG software market, serving investment firms, consulting companies, and institutional investors requiring automated analysis for portfolio decisions and regulatory compliance.

5. Ludus Automated Game Balancing: An AI-Driven Evolutionary Pipeline for Card Game Meta-Stability
Project by Rami Aldrea

This project addresses the critical and often labor-intensive challenge of balancing a complex digital card game meta, specifically focusing on a 70-card set. The primary objective was to develop and validate an automated, iterative pipeline capable of achieving and maintaining a stable, competitive game environment where no single card or archetype consistently dominates. The desired outcome was a game meta with a mean card win rate of 50%, minimal variance (ideally <5% standard deviation for competitive e-sports), and preserved strategic diversity, all while significantly reducing manual design time.

20:00 - Open for networking

20:30 - Wrap up

We have limited seat so please RSVP soon. See you all at the event.