ORTEC
Job description
At ORTEC, we solve thousands of optimization problems every day. While every problem is unique, there are patterns in the problem instances that we solve. Can we exploit the patterns in these problem instances, translating them into patterns for good solutions that we can learn from previously optimized solutions?
The challenge
Some successes have been made in using deep learning models to predict solution structures for combinatorial optimization problems, yet they are limited to simpler problems of moderate size (see some of the references below). We would like to scale up these techniques to larger and more realistic problem instances (e.g. real world routing problems with 1000s of customers). To do so, you need to develop efficient deep learning training pipelines that can run advanced (graph) neural network architectures like transformers on GPU.
Who you are
What we offer
The next steps:
Did we peak your interest? Then upload your CV, motivation letter and grade lists (BSc and MSc in one PDF if applicable). The recruitment process will consist of two online assessments, first interview, and a second interview with a case.
What to expect
We will help you to thrive in your field of expertise. We offer development programs, tailored to your individual needs and function requirements, including opportunities to attend courses and seminars. We offer challenging, practical hands-on experience with opportunities to work abroad. We operate in a flat organizational structure that keeps communication lines short. The atmosphere is open, informal, cooperative and positive.