The posting
About Boltz
Boltz is a public benefit company building the next generation of AI-powered molecular modelling tools to make biology programmable and accelerate drug discovery, while keeping frontier capabilities broadly accessible. Boltz-1, Boltz-2 and BoltzGen are open models trusted by scientists across biotech and academia, and used in programs at leading pharmaceutical, agricultural, and industrial research organisations. Our goal is to build the foundational models and molecular design systems that will transform how therapeutics are discovered and developed.
About the Role
We are looking for a Small Molecule Computational Scientist to work at the intersection of computational chemistry, cheminformatics, and machine learning.
You will support both Boltz's internal research and external drug discovery programs, providing the computational chemistry, molecular data analysis, and scientific tooling needed to answer questions across both. Internally, you will work closely with our Research and ML teams to build better datasets, evaluate our models, understand their failure modes, and guide the development of our small-molecule modelling capabilities. Externally, you will work on collaborations with pharmaceutical and biotechnology partners, applying these capabilities to real drug discovery programs and helping partners prepare, analyse, and interpret their molecular and experimental data.
You will work across small-molecule structures, protein sequences and structures, protein–ligand complexes, binding measurements, screening data, and predicted molecular properties. You will develop methods for standardising and integrating these data, analysing large molecular collections, and constructing scientifically meaningful datasets for training and evaluating our models.
A core part of the role is exploratory computational science. You might investigate the chemical diversity of a screening library, cluster millions of generated compounds, analyse model performance across protein families or chemical series, identify leakage between training and evaluation datasets, or analyse the results of an external screening campaign to understand where our models are succeeding and failing.
The role combines scientific analysis, software development, and applied computational chemistry. You will be expected not only to run existing workflows, but to understand the chemistry and biology behind the data, develop the tooling needed to interrogate it, and turn complex analyses into conclusions that can guide both our research roadmap and real drug discovery decisions.
About You
Essentials
You have a PhD or equivalent experience in computational chemistry, cheminformatics, chemical biology, structural bioinformatics, or a closely related field. You are a strong programmer, particularly in Python, and are comfortable building your own scientific tooling rather than relying exclusively on existing software. You write clean, reusable code and have experience with tools such as RDKit for manipulating and analysing chemical structures. You have a deep understanding of small-molecule chemistry and can reason about molecular structures, chemical series, functional groups, stereochemistry, protonation and tautomeric states, scaffolds, and molecular similarity. You are comfortable working with protein sequence and structural data, including protein-ligand complexes, binding sites, residue mappings, protein constructs, sequence similarity, and structural similarity. You are highly proficient at analysing large molecular datasets. You have practical experience with molecular fingerprints, similarity search, clustering, scaffold analysis, dimensionality reduction, diversity selection, sampling, and chemical-space visualisation, and understand when different approaches are scientifically appropriate. You understand the fundamentals of experimental binding and screening data, including measurements such as KD, Ki, IC50 and EC50, and can reason about assay context, uncertainty, replicates, censoring, and comparability between measurements. Strong communication skills
Nice to Have
You have experience supporting real drug discovery programs, including virtual screening, hit discovery, hit validation, hit-to-lead, or lead optimisation. You have worked with large chemical libraries or model-generated molecular datasets, including clustering, filtering, diversity selection, compound prioritisation, and analysis of chemical-space coverage. You have experience with molecular machine learning, including affinity or property prediction, virtual screening, molecular representations, or generative molecular design. You have contributed to open-source scientific software.
What We Offer
Competitive compensation and equity in a public benefit company London-based, hybrid working Direct impact: the pipeline you own is the input the model and platform depend on, used by leading discovery teams A small, high-calibre team that publishes its science Learning budget, wellbeing support and generous holiday
