TixelJobs
B
boltzvia muse

Research Scientist, Pre-training

London, ENG, GBPosted 1d ago
ResearchMid LevelFull-time

About the Role

**About Boltz**


Boltz is a public benefit company building the next generation of AI\-powered molecular modeling tools to make biology programmable and accelerate drug discovery, while keeping frontier capabilities broadly accessible.


Boltz began as an MIT spin\-out. Our founders built Boltz\-1 at MIT, and it became the first fully open\-source model to match AlphaFold 3's accuracy in predicting the 3D structure of biomolecular complexes.


We focus on two things:

* **Open foundation models.** Boltz\-1, Boltz\-2, and BoltzGen are open models trusted by 100,000\+ scientists across biotech and academia, and used in programs at every Top 20 pharma as well as leading agrichemical and industrial research organizations.
* **Boltz Lab.** Our commercial platform runs our latest models and design agents as reliable, production\-grade tools built around real chemistry and biology workflows. Teams start from a target and a hypothesis, then generate, evaluate, and rank candidate molecules, while we provide the compute, infrastructure, and collaboration layer so their scientists can move faster.


The two feed each other: open\-source adoption earns scientific credibility and community insight, and Boltz Lab turns that into real impact on our partners' discovery programs.


You can read more about our mission, research and product vision in our manifesto.

**About the role**


As a Research Scientist on Pre\-training, you will support core research on our next open foundation model, Boltz\-3\. The goal is state\-of\-the\-art structure prediction among open models, especially for antibody–antigen and small\-molecule complexes, and every model we build afterwards starts from it.


You will work closely with the Pre\-training team's research scientists and ML engineers, Research Engineering, who own training speed and cluster tooling, our Post\-training team, and domain experts in chemistry and biology.


This role is ideal for a mission\-driven technical leader who wants to push the frontier of structure prediction, then see that progress used by thousands of scientists.

**What you'll do**

* Drive research direction on Boltz\-3: new model architectures, training objectives and data strategies for structure prediction.
* Design and run rigorous large\-scale experiments, and analyze them carefully.
* Build the benchmarks that show state\-of\-the\-art antibody–antigen and small\-molecule structure prediction, on both mean and best\-of\-sample (oracle) metrics.
* Study how performance scales with inference\-time compute, and turn that into how the model is used.
* Learn from how the open model performs for the community and our partners, and turn its failure modes into the next round of research.

**About you**

### **Minimum Requirements**

* Extensive publication record in top\-tier ML or life\-science venues (e.g. NeurIPS, ICLR, ICML, Nature Methods). We're hiring at Senior or Staff level.
* Demonstrated strength in deep learning research: new architectures, rigorous experiments and careful analysis.
* Strong hands\-on PyTorch and scientific Python (NumPy, SciPy, Pandas).
* Experience training, scaling and evaluating large models, with a high bar for engineering quality, reproducibility and testing.
* Experience in an interdisciplinary environment, working with experts outside ML.

### **Preferred Requirements**

* Experience with structure prediction, protein or molecular modeling, or diffusion and generative models.
* Familiarity with the tools, data formats and workflows of structural biology, computational biology and chemistry.
* Experience using Boltz models, or open\-source contributions.


We know great people don't always tick every box. If this role excites you and you meet most of the minimum requirements, we'd like to hear from you.

**What we offer**

* Lead research on open foundation models used by 100,000\+ scientists.
* Work alongside one of the most talent\-dense teams in the field.
* Significant ownership and indepen
Share