Melodic Health GmbH
Berlin, DE · On-site · Full-time
Loading job details
Preparing the role information
Basic Facts Employment: Full-time Location: Munich / Berlin Working Language: English Starting date: ASAP if possible, but applications for later entry are also welcome. Compensation: €80-150k [depending on experience]. Negotiable for exceptional candidates. Significant equity possible. Opportunity & Role We’re building a team of exceptional engineers, scientists, and medical AI experts to accelerate the shift of healthcare from reactive treatment to proactive prevention. That is, we’re developing a best-in-class foundation model that uses a patient’s medical history to predict the risk of disease years before it would develop. Such prediction will enable doctors and patients to act before symptoms appear, so people can live healthier, longer lives. We're a team from the University of Cambridge, with backgrounds in ML and biomedical research, McKinsey, and healthtech venture capital. We think big, move quickly, and hold ourselves to high standards. Having recently raised our pre-seed round, we’re now growing our founding team. As our Founding Engineer, you’ll work directly with the founders to shape our technology and tackle challenging problems at the intersection of machine learning and medicine. You would enjoy this role if you have high agency, thrive on deeply technical work and are passionate about responsible AI in healthcare. What you’ll work on - Responsibilities Design the model. Decide what goes in and what comes out: which data we need, how longitudinal health records are represented, and how to achieve explainable outputs that show clinicians what is driving the risk and what to check next. Implement it reproducibly. Build the pipelines and experiment tracking that take the team from idea to trained model in days, make every release traceable for CE marking, and set the engineering standards. Train it. Plan and manage compute, deciding which experiments are worth running. Evaluate it. Define performance metrics that capture a model’s real value in clinical care, knowing we can’t optimise for everything. Build and maintain the benchmarks every model must pass before it reaches clinicians. Maintain it. Serve the model to clinicians and test it with them, version every release, and retrain as new data and evaluation findings come in, working with our evals engineer to fix failure modes. Experiences an ideal candidate brings Deep experience with sequence or representation learning (transformers, self-supervised pretraining), ideally on longitudinal or event data such as health records. You’ve trained AI models from scratch and managed the compute to do it, and have built training pipelines that are reproducible and that others can work in. You’ve taken a model to real users: served it, versioned releases and retrained it as data changed. Preferably you hold a PhD in ML / Engineering / Medical Informatics or similar. Published research or substantial open technical work. Healthcare domain knowledge: clinical workflows, medical terminology
Melodic Health GmbH
Posted via Arbeitnow
Apply Now opens Rozgoo's sign-up and application flow for this role. Updated today.
Apply Now