TL;DR
Tolemy is building Orbit, the control panel for the cell: a platform that helps scientists connect experimental data, biological knowledge, and AI models to understand and improve cellular systems.
We’re hiring a Founding Machine Learning Engineer to build the modelling systems behind Orbit. You’ll work on sparse, noisy, high-dimensional biological data, building models that can infer structure, quantify uncertainty, guide experiments, and become useful tools for scientists.
About Tolemy
Biology is becoming programmable, but scientists still lack the tools to understand, predict, and direct how cells behave.
At Tolemy, we’re building Orbit as a control layer for cell function: a platform that connects experimental data, biological knowledge, and AI models to build useful digital representations of cells. Our platform helps scientists move from observation to intervention.
We’re starting in therapeutic development, where better models and better experimental tools can help teams improve cell performance, manufacturing, and the development of life-saving therapies.
The goal is not just to predict what cells will do, but to help scientists understand why they behave the way they do, what to test next, and how to improve them.
If you want to build models and software that change how scientists understand cells, design experiments, and develop life-saving therapies, we’d love to meet you
Company highlights
- Recently funded and building fast. We’ve raised our pre-seed round from top European investors and are using it to build a small, exceptional technical team in Barcelona.
- A rare technical problem. This is ML for sparse, noisy, expensive-to-generate biological data, coming straight out of our own wet lab and labs of major partners.
- Real commercial pull. Orbit is being built around real customer problems, active collaborations, and experimental datasets in advanced therapies and biomanufacturing.
- Early enough to shape the foundations. You’ll help define our technical approach, infrastructure, and scientific validation culture from the beginning.
- A deeply interdisciplinary team. You’ll work directly with a team across cell biology, systems biology, software, product, and machine learning.
Life at Tolemy
We’re building an in-person team in Barcelona, working together in English. We support relocation and visa sponsorship where needed.
Our values are simple: stay curious, be kind always, and move with purpose. We care about technical depth, but also about how people work together, learn, communicate, and support each other.
We offer competitive salary, meaningful early-employee equity, flexible time off, flexible work-from-home arrangements, a learning and conference budget, high-quality equipment, and practical support to help you make Barcelona home if it’s not already.
Hard Technical Challenges
Orbit sits at the edge of machine learning, biology, and experimental science. Some of the hardest problems you’ll work on include:
Learning from sparse biology. Biological data is noisy, expensive, high-dimensional, and incomplete. How do we learn useful representations of cellular state from limited experimental data?
Building models scientists can trust. Cells contain real biological structure: metabolism, regulation, signalling, transport, and stress responses. How do we combine this knowledge with ML to build models that are predictive and biologically meaningful?
Guiding better experiments. Useful models should help scientists understand uncertainty, compare hypotheses, and decide what to test next. How do we evaluate models when there is no clean benchmark for “understanding a cell”?
What We're Looking For
We're looking for someone with strong machine learning judgment and experience building models for difficult real-world systems.
You should be able to reason from first principles about data, models, compute, uncertainty, validation, and product usefulness. You should also enjoy ambiguity, care about scientific truth, and want to build systems that help users make better decisions.
We care more about depth, judgment, and evidence of exceptional work than credentials.
You may have PhD-level training in machine learning, physics, biology, chemistry, applied mathematics, computational biology, or another systems-oriented discipline.
You should have strong foundations in one or more of: applied mathematics, statistics, optimisation, probabilistic modelling, causal inference, dynamical systems, scientific ML, or related areas.
You may also be an exceptional applied ML engineer without a PhD, with a track record of building models for complex real-world systems.
Experience with biological data is useful, but not required. What matters most is comfort with natural-world systems: messy, noisy, sparse, nonlinear, and only partially observed.
Nice to Have
- Experience with mechanistic models, hybrid ML, Bayesian methods, causal inference, dynamical systems, or uncertainty quantification.
- Experience designing benchmarks in novel or poorly defined problem spaces.
- Experience building models that move from research into production or user-facing workflows.
- Publications or open-source work in ML, computational biology, scientific computing, or related areas.
- Experience in a high-growth company, deep tech startup, or research-to-production environment.
Job Type: Full-time
Pay: 60,000.00€ - 100,000.00€ per year
Work Location: In person
