Machine Learning Lead

August 10, 2026 - Copenhagen, Denmark

Building the Next Generation Foundation Models of Biology

We want to build the next generation of foundational models of biology. Most of medicine - and most of today's biological AI - still reads human biology through DNA and RNA. We go further: into the spatial proteome, where proteins act in the native tissue context where disease occurs. This is biology that models have never been trained on, because the data has never existed at scale. That is what we set out to change.

The data is here. ResOne, our molecular intelligence platform, has already profiled over a billion pixels and hundreds of millions of peptide and protein measurements, and we project growing that by orders of magnitude. Now comes the part we built it all for: training models of biology that inform the next generation of diagnostic and prognostic biomarkers in oncology, neurology, and chronic disease.

If you are excited about working with a one-of-a-kind biological data platform and turning its data into models that make a real difference, we would love to hear from you.

The Role

As Machine Learning Lead at Resolute Bio, you will own the modeling roadmap of our data. Today our predictive models are linear models built on foundation-model embeddings; the next step is self-supervised and multimodal architectures trained on our own data - and your job is to lead that transition. This is a player-coach role: in the beginning you will lead the charge yourself, designing, training, and evaluating the first generation of serious models, while guiding our deep learning engineers and shaping how ML gets done here as the practice grows. You will join a small, international team in central Copenhagen, in mission-driven work that translates into real differences in patient care, not just publications. We are an AI-forward team: we use agentic tools daily and expect you to be fluent with them - but the judgment about what to model, whether a result can be trusted, and when a simple baseline beats a fancy architecture still comes from you.

You Will

• Own the ML roadmap - chart the path from today's linear baselines to the foundational models we train on ResOne's spatial proteomics data
• Lead from the front - design, train, and evaluate models hands-on, especially in the beginning
• Guide and mentor our deep learning engineers, setting the standards for how we build, evaluate, and ship models
• Establish rigorous evaluation practice - strong baselines, ablations, and the discipline to catch leakage, batch effects, and confounders before they become false discoveries
• Build self-supervised representations of our data and multimodal models that align modalities - connecting histology images and protein expression, with calibrated uncertainty you can act on
• Build the path from prototype to production - training infrastructure, model serving, and monitoring for models the platform depends on
• Shape what data we generate next - work with wet-lab and bioinformatics teams so that experiments are designed with the models in mind
• Communicate modeling results and their limitations clearly to scientists, leadership, and partners
  

You Have

Education & Experience

• MSc or PhD in machine learning, computer science, or a quantitative field, or equivalent experience
• 5+ years of applied ML experience, with a track record of taking models from idea to production and of guiding or mentoring other engineers

Technical Expertise

• Deep, hands-on expertise in modern deep learning with PyTorch, across architectures such as transformers, CNNs/ViTs, and self-supervised approaches
• Ruthless empirical fundamentals - evaluation design, strong baselines, error analysis, uncertainty calibration, and a nose for leakage and spurious signal in messy real-world data
• Experience training models at scale - multi-GPU workflows and large, high-dimensional datasets
• Experience serving large models in production - inference optimization, GPU serving, and monitoring
• Strong Python and software-engineering fundamentals, and fluency with AI-assisted and agentic development tools

Mindset & Skills

• You go wild on data, let evidence beat opinion, and are suspicious of results that look too good
• Player-coach - you lead by building, and you make the people around you better
• Pragmatic - you right-size the model to the problem and know when a linear baseline is the honest answer
• Driven by impact - you want models that help answer biological questions, not just top a leaderboard
  

Nice to have

• A strong competitive ML track record (e.g., Kaggle) demonstrating your ability to extract every drop of signal from a dataset
• Experience with biological or medical data - omics, computational pathology, medical imaging, or clinical outcomes
• Experience with self-supervised pretraining, pathology foundation models (e.g., UNI, Virchow), or CLIP-style multimodal alignment
• Familiarity with ML infrastructure and MLOps - experiment tracking, training pipelines, feature/data stores
• Publications, open-source contributions, or a public portfolio in relevant fields
  

Interested?

The role is on-site at our office in central Copenhagen. Please send your resume and a short cover letter to [email protected], with "ResBioML26" in the subject line - applications without the correct subject line will be filtered out. Feel free to include your salary expectations and any relevant portfolio links, such as GitHub, Kaggle, or publications.

As a company, we care about your potential, not your background. If this role excites you, do not hesitate to apply. We review applications as they come in and look forward to hearing from you!

• Want to learn more about us first? Visit resolute.bio to see what we are building.