Member of Technical Staff, Bio ML

Stealth - AI x Bio; Greater London; Híbrido; GBP 90.000 - 130.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

30 days annual leave

Private health cover

Pension contributions

Descripción de la vacante

Stealth - AI x Bio is building foundational infrastructure for data generation in AI biology. You will own development on the biological side of RL environments and agents, integrating protein models, and enabling reliable predictions backed by wet-lab validation.

You will collaborate with scientists, RL researchers, and bioinformaticians to scale experiments, fine-tune models, and advise on data needs. London or SF base with ownership and direct impact from day one.

Formación

• Hands-on experience building and training ML models for biology and understanding failure modes.
• Experience with protein language, structure and affinity models (ESM, Boltz-2, OpenFold, AlphaFold).
• Fine-tuning/post-training biological models on experimental data and validating against wet-lab measurements.
• Building RL environments, agent harnesses or tool interfaces for LLM agents.

Responsabilidades

• Develop RL environments for biological data, harness tools for agents.
• Integrate public protein models and build biological model interfaces.
• Fine-tune or post-train models on new measurements and assess data impact.
• Run scaling and data-mix experiments to optimize data requirements.
• Communicate modelling choices and results clearly to ML and scientific audiences.

Conocimientos

ML for biology

Python programming

PyTorch

Protein language models

Herramientas

JAX

OpenFold

AlphaFold

Descripción del empleo

The opportunity

We're a stealth startup in AI and bio, building infrastructure for data generation. The team is small and elite, the problems are hard, and the foundations are being laid right now. You'd help write them from the first line, with real ownership and a direct line to the founders.

About us

AI for biology has a data problem. The data that matters most doesn't exist yet, so we're building the infrastructure to produce it, with quality and traceability built in from day one. We're in stealth and heads-down on execution. We'll share more about what we're building once you've spoken with the team.

The role

You will work on the biological side of our RL environments and agents, starting with proteins. You will help build the environments themselves, working with our scientists on what goes into them, and build the harnesses through which agents use biological models: integrating the best public protein models, making them available as tools, and testing which ones help an agent and when their predictions can be trusted.

Where existing models fall short on our data, you will fine-tune or post-train them on our measurements. As new measurements come in, you will run scaling and data-mix experiments to establish what kind and amount of data improves the models, and turn the results into recommendations on what we measure next.

You will report into the intelligence team and work closely with the Applied RL researchers, the protein bioinformaticians, the evaluation and infrastructure engineers, and our scientists.

Who you are

You have a strong background in machine learning for biology. You have built models yourself, adapted existing ones and tested the results against wet-lab data, and you understand how a model can perform well on a held-out split and still fail in the lab.

MUST HAVE

• Hands-on experience building and training machine learning models for biology, and an understanding of where they fail.
• Experience with protein language, structure and affinity models, such as ESM C, Boltz-2, OpenFold3 or AlphaFold 3.
• Fine-tuning or post-training pretrained biological models on experimental data, and validating the results against wet-lab measurements.
• Building RL environments, agent harnesses or tool interfaces that let LLM agents use machine learning models.
• Working with noisy experimental data, including replicates, controls and batch effects.
• Running experiments on the data itself, such as scaling studies, data-mix ablations and learning curves on small datasets.
• Clear writing and communication about modelling choices, results and their limitations, for both ML and scientific audiences.
• Strong Python and PyTorch or JAX, and daily use of coding agents.

NICE TO HAVE

• Experience post-training LLM agents or protein models with reinforcement learning, for example with ProtRL.
• Experience with antibody or binder design, including working with developability, affinity or stability data.
• Experience with biological data beyond proteins, such as functional genomics or single-cell data.
• Publications or open-source work in biological machine learning.

How we work

The role is based in London or San Francisco. We work in the open, write decisions down, and expect each person to own their area and make the calls inside it. Everyone gets 30 days of annual leave plus local public holidays. In the UK that comes with a pension with a 10% employer contribution and Bupa private health cover; in the US, a 401(k) with matching and health cover. We add to these as the team grows.

Our process

Four stages: a screening call, a behavioural interview, an in-person technical session with the team on a real problem from our work, and references.

We are an equal opportunity employer. We make hiring decisions on merit, scope-fit, and the strength of the working relationship we expect to build with each hire. Applications welcome from candidates of any background.