Webseite Fraunhofer-Institut für Produktionstechnologie IPT

The Fraunhofer IPT uses machine learning (ML) models that can predict process states or product quality in production. In order to implement the results into the infrastructure of the production environment, DevOps for ML or MLOps is applied. In this thesis, MLOps concepts and methods are to be analyzed and subsequently implemented for a specific use case. The goal is to cover the whole process from initial model training to deployment, monitoring and retraining against new data.

Your tasks

  • Familiarisation with the state of the art of the deployment of ML models and MLOps
  • Analysis of existing concepts and methods of MLOps
    Deployment of an ML model in the existing infrastructure of a use case from production
  • Evaluation and documentation of the results

Our requirements

  • You are studying engineering, computer science, operations research or a comparable subject
  • You have initial experience in the field of machine learning or DevOps
  • A high degree of initiative and team spirit
  • Very good language skills in English

Your benefits

  • Pioneering role in the deployment of ML models in production
    Excellent equipment and infrastructure (cloud platform VFK and computing cluster)
  • Collaboration in a dedicated team consisting of engineers, data scientists and software developers with many years of experience in relevant fields
  • Available data basis and ML models

We look forward to receiving your application
Hendrik Mende, M.Sc.
Fraunhofer Institute for Production Technology IPT
Steinbachstraße 17, 52074 Aachen, Germany

Um sich für diesen Job zu bewerben, sende deine Unterlagen per E-Mail an hendrik.mende@ipt.fraunhofer.de