Sentaro is building a digital immune system: one that learns what's normal and stops what's not.

We are a small team of top-tier engineers looking for a Founding ML Engineer with a security mindset to own the models at the core of that immune system.

You will train and evaluate those models on LUMI, the EuroHPC supercomputer in Finland, with compute at a scale few startups can offer.

What you will do

• Take threat-detection models from raw data to production, and own every step

• Run large-scale training on LUMI end-to-end, from hyperparameter optimization to analyzing studies

• Data pipeline engineering for model training and calibration

• Own evaluation and model releases, adversarial robustness included

• Ship models into a Go inference service. Work with backend on serving, rollout and rollback

• Turn messy real-world threat data into explainable verdicts with our security experts

What we are looking for

• Several years shipping ML models in production, not just notebooks

• Hands-on LLM fine-tuning and evaluation, or large-scale distributed training

• Strong Python skills and experience with the modern stack: PyTorch, Hugging Face, experiment tracking and hyperparameter tooling

• Enough Go, or willingness to learn it, to reason about how your model runs in production

• Curiosity about attacker behaviour and adversarial AI in order to turn those into features

• On site with the team in Stockholm

What we offer

• Significant equity. You join early and own a real piece of what you build

• A small team of top-tier engineers

• Compute most startups can only dream of, through our access to LUMI

• Freedom to take research ideas to production at a scale where they become reality