Senior ML Engineer | Data Scientist
Summary
Own motility models end to end at Digestaid: translate clinical questions into time-series and signal-processing ML problems, ship production-grade models through a regulated release process, mentor data scientists, and contribute to regulatory/scientific documentation. Stack: Python, PyTorch, scikit-learn, GitHub Actions, GCP (Vertex AI, Cloud Build).
Own our motility models from problem framing to production deployment, turning clinical questions into time-series and signal-processing problems.
Motility models, end to end
Own the design, training, validation, and maintenance of motility models, from problem framing to production deployment.
Clinical questions into ML problems
Turn clinical questions into well-posed time-series and signal-processing problems.
Set and defend the evaluation strategies.
Production-grade code
Ship production-grade code with testing, logging, and reproducibility to our regulated release process.
Mentor data scientists on the team.
Regulatory and scientific writing
Contribute to technical documentation supporting regulatory submissions and to scientific publications where relevant.
What we're looking for
Models that reach production
and stay there.
Requirements
- MSc or PhD in Computer Science, Software Engineering, Biomedical Engineering, or similar
- 5+ years of applied experience, with a track record of ML models that reached production and stayed there
- Strong grounding in time-series and signal processing
- Fluent Python, solid data structures and algorithms, and good software engineering practice
- Experience working with coding agents as part of daily engineering work
- Comfortable working in a regulated environment: version control discipline, traceability, and writing things down
- Good English communication, including with non-technical clinical stakeholders
Nice to have
- PhD in machine learning, signal processing, or biosignal analysis
- Experience with physiological or biomedical time-series
- Experience building agentic or LLM-backed internal tooling
- Experience with medical device software (IEC 62304, ISO 13485) or another regulated domain
- Data visualization for clinical and scientific audiences
- Working Portuguese (not required, our working language is English)
PyTorch scikit-learn GitHub Actions Google Cloud: Vertex AI, Cloud Build

