Principal Scientist, AI Co-Scientists – Synthetics Development Full-time NEW

Principal Scientist, AI Co-Scientists – Synthetics Development

The future of small-molecule discovery and development depends on both translating scientific challenges into trustworthy AI-enabled capabilities and creating the data, information, and knowledge foundations on which those capabilities depend. This role will help Synthetics Development build and evaluate practical solutions spanning data science, machine learning, generative AI, information retrieval, document understanding, structured knowledge, and agentic workflows.

In this role, you will work at the interface of small-molecule science, data science, and artificial intelligence. In close partnership with chemists, analytical scientists, formulation scientists, process engineers, data scientists, and technology teams, you will investigate, prototype, and evaluate AI-enabled capabilities that address priority scientific challenges.

The role combines two closely connected responsibilities. You will de-risk new ideas through hands-on experimentation, focused technical spikes, and limited-scope proofs of concept. You will also design and implement the document-processing, metadata, retrieval, structured-data, ontology, and knowledge-graph components required to make scientific information usable by these solutions. In both areas, you will translate scientific needs into testable technical approaches and make feasibility, limitations, dependencies, risks, and potential value visible before broader investment.

Working as part of a cross-functional team, this role will contribute to scientific co-scientists, intelligent search, scientific data products, workflow automation, decision-support tools, and agentic workflows. The work will span practical experimentation and the implementation of reusable information foundations, including ingestion and document understanding, metadata and entity extraction, retrieval, controlled vocabularies, ontologies, and knowledge graphs. The role is not responsible for foundation-model development or production platform ownership. Its focus is hands-on development, technical learning, and creating robust components and evidence that help the team decide what to build and scale.

This is a rare opportunity to help shape AI-native Synthetics Development by combining direct experimentation on real scientific challenges with the hands-on construction of the information and knowledge foundations required for credible scientific AI. You will work with scientists to understand their workflows, build and test targeted solutions, and ensure that the underlying scientific content can be processed, structured, retrieved, and reused effectively.

The resulting prototypes, ingestion and retrieval pipelines, structured knowledge assets, reusable components, experimental findings, and technical recommendations will inform capabilities across small-molecule discovery and development, including chemistry, formulation, process engineering, analytical sciences, and increasingly lab-integrated workflows.

Key Responsibilities

• Applied AI Prototyping & Technical De-risking
• Design, implement, and evaluate limited-scope proofs of concept for priority scientific use cases.
• Run focused technical spikes to test assumptions, compare approaches, and assess emerging AI, machine-learning, and scientific-computing technologies.
• Translate scientific needs into testable hypotheses, representative experiments, success criteria, and clear technical questions.
• Rapidly build prototypes that make feasibility, limitations, dependencies, risks, and potential value visible before broader investment.
• Document experimental setup, results, trade-offs, and recommendations so that findings are reproducible and actionable.
• Identify when an approach should be refined, handed over for engineering and scale-up, or discontinued.
• Scientific AI & Data Science
• Apply data science, machine learning, generative AI, information retrieval, and agentic approaches to scientific challenges in small-molecule discovery and development.
• Build experimental workflows that combine scientific data, documents, models, knowledge sources, APIs, and specialist tools.
• Investigate the behavior of AI-enabled workflows, including the quality of outputs, evidence use, tool interactions, failure modes, and sensitivity to input or configuration choices.
• Contribute reusable code, components, datasets, and experimental patterns that accelerate subsequent scientific AI work.
• Use quantitative and qualitative evidence to communicate what works, what does not, and what is required to progress a concept.
• Make scientific data AI-ready and science-ready by structuring, contextualizing, and curating experimental, analytical, process, molecular, and modeling data so it can be consumed by scientists, predictive models, co-scientists, and agents.
• Assess the AI-readiness of candidate data sources, identifying gaps in quality, completeness, context, and accessibility, and prioritize data improvements that are driven by concrete scientific use cases rather than completeness for its own sake.
• Scientific Information, Retrieval & Knowledge Foundations
• Work with subject-matter experts to understand how scientists find, interpret, combine, and use information in their work.
• Design and test approaches for scientific search, retrieval, question answering, and information reuse.
• Contribute to document processing, metadata, structured representations, controlled vocabularies, ontologies, or knowledge graphs, applying FAIR (findable, accessible, interoperable, reusable) principles pragmatically when these are needed to solve the scientific problem.
• Evaluate combinations of semantic search, keyword search, filtering, aggregation, structured queries, and knowledge-based methods.
• Help diagnose whether failures arise from data quality, missing knowledge, retrieval, reasoning, tool use, orchestration, or unclear requirements.
• Build reusable, well-documented data pipelines and data products, including ingestion, transformation, and contextualization, that connect data across laboratory, process, and scientific systems and make it accessible through APIs for AI, analytics, and automated workflows.
• Partner with data, architecture, and IT teams to leverage existing enterprise data platforms rather than recreating them locally, and provide the scientific perspective on semantics, provenance, and access.
• Cross-Functional Delivery & Continuous Learning
• Act as a practical bridge between scientific SMEs and AI, data science, product, architecture, and engineering teams.
• Support co-design sessions, pilots, user testing, and technical demonstrations with scientific communities.
• Collaborate with engineering and platform teams to define the requirements and boundaries for solutions that are selected for further development.
• Contribute to evaluation, provenance, traceability, documentation, and governance appropriate to scientific and quality-conscious environments.
• Track relevant advances in AI, machine learning, scientific computing, retrieval, and agentic systems, and assess them through targeted experimentation rather than technology-led adoption.

Ideal Candidate Profile

The ideal candidate is a scientifically curious, pragmatic, and hands-on collaborator who enjoys turning ambiguous scientific challenges into concrete experiments. They can move quickly from a question to a working prototype, while maintaining enough rigor to distinguish a promising capability from a plausible-looking demonstration.

They combine strong data-science and software-development skills with scientific intuition, critical thinking, and clear communication. They are comfortable working between subject-matter experts and technical teams, challenging assumptions, learning from failed approaches, and balancing rapid exploration with reproducibility, maintainability, and responsible use.

Qualifications

Required

• Advanced degree in data science, computer science, bioinformatics, cheminformatics, computational science, pharmaceutical sciences, chemistry, engineering, or a related field; or equivalent relevant experience.
• 4+ years of relevant experience in applied data science, machine learning, scientific informatics, scientific software, pharmaceutical research and development, or a closely related environment.
• Hands-on proficiency in Python, including the ability to implement, test, debug, and maintain experimental AI and data-science workflows.
• Experience designing and executing technical experiments, prototypes, proofs of concept, or feasibility studies.
• Experience developing or evaluating AI-enabled scientific applications, co-scientists, decision-support tools, or tool-using agents.
• Experience with retrieval-augmented generation, semantic or hybrid search, vector databases, knowledge graphs, scientific NLP, or document intelligence.
• Experience working hands-on with scientific data, scientific documents, models, or scientific information systems, including data or document processing, metadata, retrieval, or structured knowledge.
• Familiarity with scientific data sources, predictive models, cheminformatics tools, simulation environments, experiment-design methods, or laboratory systems.
• Working knowledge of Git and reproducible software-development practices, with familiarity with Linux or shell environments, APIs, containers, and automated testing.
• Demonstrated ability to translate scientific needs into technical hypotheses, experiments, prototypes, and evidence-based recommendations.
• Strong analytical and communication skills, with the ability to collaborate across scientific and technical disciplines and explain complex trade-offs clearly.
• A collaborative, iterative working style and the ability to manage ambiguity while maintaining scientific rigor, traceability, and focus.

Preferred

• Hands-on experience in small-molecule discovery or development, such as synthetic chemistry, analytical development, formulation, process development, high-