Machine Learning / Computational Scientist (Same-Cell Biology) 80-100% (m/f/d)
Glattbrugg,
Switzerland
Measuring a cell has always meant destroying it, which is why most AI in biology learns patterns rather than causes. FluidFM, our core technology, removes that constraint. Live-seq, our non-destructive sampling method, lets us read a living cell without killing it, so we can measure the same cell before and after a perturbation, across time, transcriptome and phenotype. In a market where ~90% of clinical trials fail and an approved drug costs billions, knowing what causes a response rather than what correlates with it determines whether a decade of development is well spent.
Cytosurge is building a team to leverage Live-seq in the world of cell prediction models, building models that quantify mechanisms rather than fit correlations, from a rare kind of data. As our Machine Learning / Computational Scientist, you own the computational models at the scientific core of the program, reporting to the Head of Causal Bio Program. You are a key figure in carrying the program through its milestones, from first proof of concept to a working platform.
Your Challenge
- Own the models end to end: Design, build and validate the models that turn paired same-cell data into causal insight, including the evaluations and baselines that decide whether the approach works
- Critically assess the cell-prediction landscape: Which existing methods transfer well to same-cell paired data, which break, and how to exploit the unique properties of Live-seq
- Build our in-silico engine to drive mechanistic understanding, and to decide which experiment is worth running next
- Own the pipeline’s development and maintenance cycle, and shape its architecture so it scales as the program grows
- Work at the hub of a small cross-functional team (molecular biology, lab and data generation, engineering, product): Turn biological questions into testable causal ones, and your results into direction on where the program goes next and which wet-lab work to prioritize
Your Profile
- MSc or PhD in machine learning, statistics, mathematics, computational biology, physics or a related quantitative field, or equivalent research experience gained in industry, with depth in causal and statistical ML, or mechanistic modelling of biological systems
- You come from the world of machine learning and computational modelling applied to biology and cell prediction models: Perturbation-response prediction, pathway modelling, virtual-cell or cell-state modelling, gene-regulatory-network inference, or comparable
- Hands-on experience with single-cell omics data, multi-modal ML, and a critical view of what today's computational methods can and cannot deliver in the context of cell prediction
- Comfortable with complex, noisy data (temporal, transcriptomic and phenotypic) in the small-sample regime, where a few well-designed measurements matter more than many unpaired ones, on a solid mathematical and statistical foundation (experimental design, identifiability, power, uncertainty)
- Builder: fluent in the Python ML stack, you write clean code from scratch and take a research idea all the way to a working, well-validated model, in small increments rather than one big build
- ML system and architecture design: You see how components and data fit together, and build so the work scales cleanly as the program grows
- Rigorous, self-critical and independent: You design sound evaluations, know when to trust a result and when to question it, and are comfortable with ambiguity in a research-phase program
- A strong communicator in both directions: You explain technical choices and results so that molecular biologists, engineers and product colleagues can act on them, and you draw out what you need from them in return
- Business fluency in English is required
Our Offer
- Ownership of the scientific core of a strategic program, with the freedom to shape the technical foundation of the prediction stack, and direct influence on where experimental capacity goes
- The chance to push the boundary of cell predictive models, on a kind of data no one else has, produced in-house with the lab team next door rather than downloaded from a public archive
- Visible impact, as the model capabilities you build feed directly into our solutions
- A small, agile, cross-functional team where the path from question to experiment is short and direct
- A structured goal setting and execution framework based on Objectives and Key Results (OKRs)
- Opportunity to visit relevant conferences to remain up to date and on top of technology developments
- An innovative, collaborative work environment where your ideas are welcomed and valued
- High potential for professional growth in a fast-moving biotech company
Why should you join our team?
As our Machine Learning / Computational Scientist you build the models that exploit what Live-seq makes possible, and push the boundary of cell prediction. You get the chance to work on something that no one else has done before. We are looking forward to discussing this adventure and meeting you in person!
