Why this role exists
Cancilico is a Dresden-based, seed-funded health-tech company, spun out of TU Dresden and University Hospital Dresden. Our first product, MyeloAID, uses AI to automate the morphological differentiation of bone marrow smears. No one else in the EU is doing AI-driven bone marrow analysis at this depth, and we’re expanding into new diagnostic verticals (peripheral blood, and beyond).
Every new vertical we open comes down to one loop: get the right data → train a model fast → evaluate it rigorously → report it → iterate. This role owns that loop. Your evaluations will convince clinicians and regulators when we make claims, and they’re our path to CE-IVDR clearance.
What you’ll do
- Build and own automated data pipelines — ingestion, cleaning, versioning, and QA of whole-slide images, tiles, and annotations across hospitals, scanners, and stains. Extend our internal data registry as new verticals demand new stratifications.
- Stand up MVP models fast for new verticals (detection, classification, regression), then improve them deliberately.
- Own our evaluation framework — the metrics that tell us whether a model is good enough to ship and good enough to defend to clinicians and regulators. Inter-rater benchmarking, drift tracking across scanners, cameras and slide preparation variations.
What we’re looking for
Must have
- Strong applied ML/data engineering — you’ve taken models from data to evaluated result, not just trained on clean benchmarks.
- Solid Python and data-pipeline tooling; comfort with experiment tracking and reproducibility.
- A rigorous, skeptical attitude to evaluation — you instinctively ask “how would this metric mislead us?”
- You’re fluent with modern AI coding tools (Cursor, Claude Code, similar) and use them daily as part of your default workflow.
Nice to have
- Hands-on PyTorch experience for vision tasks — classification, object detection on COCO-style annotations, or regression.
- Hands-on experience with DVC, Prefect, MLflow, or comparable data-versioning / orchestration / experiment-tracking / serving stacks.
- Experience with imaging / computer-vision data, ideally medical or microscopy (whole-slide images, tiling, stain normalisation).
- Exposure to regulated environments (IVDR, MDR, ISO, GxP) or a willingness to learn them.
- Familiarity with annotation workflows and inter-rater quality.
You do not need a medical background, but you should be genuinely curious about the clinical problem.
Our stack & setup
- Data: PostgreSQL + DVC + Prefect, orchestrated through our internal data registry.
- ML: PyTorch + torchvision, MLflow for experiments, FastAPI for serving.
- Tooling: GitHub for code, CI/CD, and tasks.
You’ll report to our CAIO, Sebastian, and work alongside an AI engineer, an AI intern, and ~10 clinical annotation consultants (hematologists and medical technical assistants).
What we offer
- Salary: €70,000–€95,000 depending on experience and fit.
- 30 days paid leave, flexible core hours (10:00–15:00).
- 40-hour weeks.
- Pick your machine: Mac, Windows, or Ubuntu.
- Flexible German benefits, negotiable based on your needs.
- Working language is English. No German required.
- Hematologists and medical technical assistants are colleagues you can walk over and ask.
- Hybrid in Dresden preferred. Fully remote within EU time zones possible for the right candidate. You’ll need an existing right to work in the EU; we can’t sponsor relocation or visas at this stage.
- We welcome applications regardless of background, parental status, disability, or neurodivergence.