Data Discovery: You start your journey by conducting a deep dive into our data, mapping telemetry, service tickets, and machine metadata to build a labeled dataset linking sensor signals to historical failure events
Method Benchmarking: You perform a rigorous method benchmarking by comparing classical anomaly detection methods (e.g., Isolation Forest, ARIMA) against modern time-series deep learning approaches (e.g., LSTMs, Transformers)
Proof of Concept: You implement a robust proof of concept, applying your best model to specific machine classes to simulate the alert-to-ticket workflow and measure early breakdown detection
Business Case Quantification: You quantify the concrete business case, analyzing cost reduction potentials, revenue upside opportunities in service contracting, and future product enhancements to capture additional value
Academic Write-up & Publication: You are responsible for the evaluation, interpretation, and documentation of the results in English, preparing both your thesis and a co-authored research paper for publication in reputed journal

Educational Background: You are about to complete a Master's degree in Computer Science, Data Science, Statistics, Industrial Engineering, or a comparable course of study, and you would like to write your thesis within six months
Technical Expertise: You have strong foundations in time-series modeling and anomaly detection, with proficiency in Python (pandas, scikit-learn, PyTorch/TensorFlow) and SQL. Exposure to Snowflake or Google Cloud is a plus
Bridge Builder Mindset: You love crossing the chasm between theory and practice, translating complex model outputs into tangible business impact at the boundary of research and applied engineering
Working Style & Ambition: You work in a structured, independent manner, bring a high degree of initiative, and have the clear ambition to publish your results in a peer-reviewed paper
Language Proficiency: You have very good English skills (as the thesis and publication will be written entirely in English)

Culture : Be part of a modern company culture where your work is designed from day one to be co-authored and published in reputed international journals
Top-Tier Mentorship: Receive direct, high-impact supervision from our VP Customer Support and Service, our AI Transformation Lead, and your academic supervisor
Flexible Workspace : Benefit from a highly flexible setup: while Munich is preferred, remote candidates are welcome, and travel costs to Munich for working sessions will be fully covered
Team Integration: Enjoy a deeply embedded collaboration within EGYM’s AI Transformation and Service Operations teams, combining cutting-edge tech with practical execution
Real-World Data: Gain access to a unique, massive industrial dataset spanning thousands of connected smart fitness machines globally
Fitness : Train in our inhouse gym with all EGYM Smart Strength and Smart Flex machines for free in case you are completing your master thesis in Munich

Master Thesis Student – Predictive Maintenance for Connected Fitness Equipment

Key skills