Senior MLOps Engineer
Description
We're looking for Senior MLOps Engineer for our client in Stockhlm.
The Position
You'll be working with a rapidly growing deep-tech startup building a platform for the manufacturing industry and apply ML/AI to how physical products are designed and produced. The ML team is shipping, and the GPU platform underneath it now needs a dedicated owner. You’ll take the training and evaluation infrastructure that already works and decide how it scales, so the ML engineers can focus on models rather than machines. This is a high-ownership role with very little process between you and the decisions.
What you’ll work on
- Training infrastructure across on-prem GPUs in the office and cloud capacity on AWS
- End-to-end reproducibility: versioned datasets, tracked experiments, and every model traceable back to the data it learned from
- The infrastructure behind their generative design agents and the evaluations that test them
- Interactive GPU environments that pause when idle, keep their state, and resume within seconds on any machine
- Keeping GPU spend under control and customer data secure
- Working directly alongside the ML engineers and clearing whatever slows them down
Requirements
- 5+ years experience building and running ML infrastructure
- Comfortable working directly with GPUs, from driver and CUDA issues through to finding capacity and keeping cloud bills sensible
- IaC with Terraform/Terragrunt and strong cloud skills (Ideally AWS)
- You get uneasy when an experiment can’t be rerun or data isn’t properly locked down, and you build systems that prevent both
- Fluent English speaker, able to work onsite in Stockholm.
Nice to have: some time on the modelling side yourself, or experience running LLM-based agents for real users
The offer
- Above market Base salary
- Attractive Stock option plan
- ITP Pension Plan
- A small, skilled but humble team that writes things down, proposes before it builds, and measures before it claims
- Real autonomy and trust from day one
Skills
Sign in to read and join the discussion.