MLOps: DevOps for Machine Learning – A State of the Art Research

Bachelor thesis, Master thesis

Research focus: Development Tools, Industrial Automation, Machine Learning
Courses of study: Informatics, Information Management, Mechanical Engineering, Related Disciplines, Business Informatics, Industrial Engineering and Management

Job Description

The FZI Research Center for Information Technology is an practice-oriented research institution in Baden-Württemberg with close contact to industry. The research focus of our department lies on novel methods for industrial automation and intelligent solutions based on Artificial Intelligence.

According to the Harvard Business Review magazine, Data Scientist is the "Sexiest Job" of the 21st century. Continuously increasing numbers of assocaited jobs in companies of various sizes, seem to reflect this claim indeed. Nevertheless, the development of approaches based on Artificial Intelligence (AI) does not seem to reliably translate into productive solutions. It is often observed that development cycles for performance optimization of AI approaches take up plenty of time and do not leave the experimental stage.
A currently frequently pursued approach to combat this is found in DevOps. This approach, borrowed from software development and engaging at the level of corporate mentality, includes the practice of CI/CD (Continuous Integration / Continous Delivery). Transferred to the development of AI and Machine-Learning-based approaches, the term MLOps is commonly used. MLOps describe procedures that serve the efficient, smooth and automated conceptual design, development, testing and deployment of AI-based solutions.

Your Responsibilities

  • Research and formulation of the state of the art in the field of MLOps
  • Compilation of important terminology, methods and available solutions
  • For Bachelor Thesis: Extension with the possibility to bring in own ideas

Our Offer

  • An interdisciplinary environment amidst science, business and transfer into practice
  • A welcoming atmosphere in a young and motivated research team

Your Profile

  • No substantial programming skills required
  • Ideally, initial practical experience in Data Science, Artificial Intelligence, Machine Learning (e.g. TensorFlow, Scikit-Learn)
  • Optionally: Experience in the field of Docker, Git, Software Testing, REST
  • Fluent in German or English
  • Highly motivated students able to work proactively and independently


Please attach the following to your application and send it to Martin Trat,

  • Motivation for this application condensed into two-three sentences
  • Transcript of records / grades

Job Description

  • Start: with immediate effect
  • Can be submitted as Seminar Paper oder Bachelor Thesis
  • Associated KIT institute: Institut für Informationsmanagement im Ingenieurwesen (IMI), Prof. Dr. Dr.-Ing. Dr. h. c. Jivka Ovtcharova
  • Succeeding offer to work as scientific assistant in our team possible