Principal ML Scientist – Predictive Toxicology

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Principal ML Scientist – Predictive Toxicology based in the United Kingdom.

This role offers the opportunity to shape the future of machine learning applications in life sciences and drug discovery.
You will lead the scientific strategy behind predictive toxicology and quantitative biology initiatives, transforming complex biological data into impactful AI-driven solutions.
Working at the intersection of machine learning, computational biology, and pharmaceutical research, you will help develop models that improve therapeutic discovery and decision-making.
The position combines scientific leadership with hands‑on technical contribution, allowing you to influence both product direction and customer outcomes.
You will collaborate with industry partners, researchers, and technical teams to integrate advanced modelling approaches into real‑world workflows.
This is a high‑impact opportunity for a scientist who wants autonomy, ownership, and the chance to advance AI‑powered innovation in healthcare.

Accountabilities:

The Principal ML Scientist will own the expansion of predictive toxicology and quantitative biology capabilities, defining scientific direction and delivering machine learning solutions that create value for life sciences partners.

• Lead the development and execution of the scientific strategy for predictive toxicology, quantitative biology, and related drug discovery workflows.
• Define modelling approaches, biological endpoints, and data strategies that support better safety and efficacy decisions in pharmaceutical research.
• Build and optimize machine learning models using advanced molecular AI techniques, including approaches such as graph neural networks, message‑passing architectures, and transformer‑based models.
• Apply federated learning approaches to enable collaborative model development across multiple organizations while maintaining data privacy and ownership.
• Integrate scientific workflows involving areas such as multi‑omics, image‑based screening, high‑throughput screening, and compound prioritization into scalable solutions.
• Collaborate directly with customers and scientific partners, leading discussions around evaluation, adoption, delivery, and roadmap development.
• Translate complex scientific challenges into practical AI solutions that can be incorporated into real drug discovery programs.
• Mentor other scientists and contribute to building future scientific capabilities within the organization.

The ideal candidate combines deep expertise in machine learning applied to life sciences with strong scientific leadership and the ability to work independently across technical and customer‑facing environments.

• PhD or equivalent experience in computational biology, cheminformatics, toxicology, machine learning, or a related scientific discipline.
• 6+ years of experience applying machine learning techniques to drug discovery, computational biology, or life science challenges.
• Strong understanding of deep learning methods for molecular AI and predictive modelling.
• Proven experience developing predictive toxicity models and supporting their adoption within pharmaceutical or industrial research environments.
• Knowledge of toxicity assessment workflows, including areas such as DILI, cytotoxicity, genotoxicity, or related safety endpoints.
• Experience working with biological datasets such as RNA‑seq, toxicity screening data, image‑based screening, or high‑throughput screening workflows.
• Ability to define scientific vision, lead technical discussions, and communicate effectively with customers, partners, and internal teams.
• Strong hands‑on modelling skills combined with the ability to guide scientific strategy and mentor others.
• Excellent analytical, problem‑solving, and communication skills.
• Professional working proficiency in English.

Nice‑to‑have qualifications:

• Experience with federated learning, privacy‑preserving machine learning, or distributed AI systems.
• Experience validating predictive toxicity models prospectively and influencing compound design or prioritization…