- Where
- Munich, Germany · Gothenburg, Sweden
- Level
- Mid
Would you like to help deep learning engineers spend more time solving meaningful problems and less time managing complex workflows? Are you interested in automation, evaluation, observability, and the practical use of agentic AI? In this role, you will help shape how deep learning development works across Zenseact. You will join a team that connects deep learning development with reliable operations. Your work will support a continuous flow of trained and evaluated models, while giving engineers and decision-makers clearer information about performance. Apply today, no CV needed at first stage, only contact information, we'll cover the rest later.
Insights from the team
You will join team DLFlow within our MLOps unit. Our MLOps teams create the systems and workflows that bring deep learning development together, enabling the end-to-end deep learning development; including training, evaluating, and deploying hardware-optimized models into the product code. Zenseact’s deep learning teams explore what needs to be solved and develop new ways to solve it. Our team helps turn those advances into a smooth, repeatable process for training, evaluating, and integrating models. We work closely with deep learning developers and other stakeholders across Zenseact to create an efficient AI factory.
The work sits at the intersection of deep learning and operations. We draw on ideas from CI/CD but adapt them to the needs of machine learning. Our focus is to make development more efficient and help deep learning engineers spend most of their time on deep learning itself. Automation, evaluation, and observability are central to what we deliver.
Your role and day-to-day tasks
Your role is to improve the way deep learning models move through development, evaluation, and delivery. You will create reliable workflows that make complex work easier to repeat, understand, and improve. You will also help teams decide which metrics matter and present results in a way that supports sound decisions.
A key part of the role is pushing the boundaries of how agentic AI can be integrated into the development cycle itself, rather than remaining a separate tool developers invoke as needed. You will explore new approaches, test them in practice, assess where they create real value, and turn the most promising ideas into dependable working methods.
Typical tasks include:
Build and improve automated workflows for training, testing, and evaluating deep learning models.
Develop observability tools that make model performance easier to understand.
Turn large sets of metrics into clear insights for engineers and decisionmakers.
Explore how agentic AI can improve deep learning development processes.
Apply CI/CD principles to deep learning and machine learning workflows.
Work with deep learning teams to understand their routines, pain points, and technical needs.
Discuss priorities and solutions with stakeholders, then help establish sustainable ways of working.
What’s in it for you
You will work on an engineering challenge that does not come with a ready-made answer. You will have room to bring in lessons from other environments while learning from people who work directly with deep learning every day.
Your contribution will give other engineers more space to focus on their core work. At the same time, you will have the opportunity to help shape how new technology, including agentic AI, is introduced thoughtfully into development processes. You will work across team boundaries and help ideas become reliable practices.
Qualifications & experience
We believe you have practical knowledge of deep learning and understand what a typical development cycle looks like. You know the common pain points and can communicate effectively with deep learning engineers. Your interest goes beyond creating model improvements yourself. You are motivated by building the systems and procedures that help an entire development organization work effectively.
You follow new developments in deep…