Senior Data Scientist, Agentic AI & LLM Engineering

CAS•Columbus, OH

3d•Onsite

About The Position

CAS is seeking a Senior Data Scientist focused on agentic AI development and large language model (LLM) engineering. This role is part of the Data Analytics and Insights (DAI) team, which builds the AI that powers CAS's scientific information products. The position involves working on systems like the Newton research assistants, natural-language query agents, predictive models, and AI-assisted content curation pipelines. The ideal candidate is a highly self-directed professional capable of independently designing, building, and evaluating agentic AI systems and LLM-powered applications. This role requires providing strategic input into product and platform ideation, driving complex technical initiatives from concept to production, and translating advanced AI engineering into tangible business value with minimal oversight, while also providing technical direction to less-experienced team members.

Requirements

Master's degree in Computer Science, Data Science, Chemistry, Materials Science, Chemical Engineering, or a related technical or scientific field with 6+ years of applied AI/ML or data science experience; a Bachelor's degree in one of those fields with 8+ years; or a PhD with 3+ years.
Advanced proficiency in Python and the modern AI engineering stack (for example, PyTorch, Hugging Face, vector databases, orchestration/agent frameworks, SQL, and cloud platforms).
Demonstrated, hands-on experience building LLM-powered applications, agentic or multi-agent systems, and/or RAG pipelines — and taking them to real users in production, including the post-launch iteration that implies.
Experience fine-tuning, adapting, or deploying large language models, including familiarity with self-hosted or open-weight models.
Working knowledge of evaluation methodology for generative and agentic systems (accuracy, faithfulness, and hallucination/variability management).
Ability to work with scientific data and domain concepts — such as chemical or materials data, molecular representations, or life-sciences datasets — well enough to judge whether a system's outputs are accurate, defensible, and useful to working scientists, and to partner effectively with subject-matter experts.
Exceptional communication skills, with a proven ability to influence senior stakeholders and lead cross-functional efforts.
Strong project-management capability across multiple concurrent initiatives.
Experience deploying and operating production systems in at least one cloud environment (AWS, Azure, or GCP).
Proficiency with agentic software development tools (for example, Claude Code or similar AI-assisted development environments).

Nice To Haves

PhD in a computational, physical, chemical, or life-sciences field.
Experience with tool-use and integration frameworks such as the Model Context Protocol (MCP).
Experience self-hosting or serving open-weight models in production (LLMOps / MLOps).
Cheminformatics experience or familiarity with molecular representations and scientific databases.
Experience building and deploying predictive ML models in a scientific or regulated domain.
Track record of publications or conference presentations in AI-for-science or applied machine learning.
Consulting or client-facing experience in scientific, chemical, or life-sciences sectors.

Responsibilities

Independently design, build, and productionize agentic AI systems — including multi-step, tool-using, and multi-agent workflows — that support scientific research tasks.
Develop retrieval-augmented generation (RAG) pipelines that ground model outputs in curated, provenanced scientific content.
Implement and extend model-context and tool-integration frameworks (for example, the Model Context Protocol) to connect LLMs and agents to scientific data sources and external systems.
Apply prompt engineering, context engineering, and orchestration techniques to improve the reliability and quality of agentic and generative workflows.
Fine-tune, adapt, and deploy large language models — including self-hosted, open-weight models — for domain-specific scientific tasks.
Develop and deploy machine learning models that ship inside CAS products, such as property, toxicity, or biologic developability prediction, using scientific judgment to inform feature design and validate model behavior.
Take AI capabilities from prototype all the way through production: building the data, training, and inference pipelines, owning the deployment, and iterating after first release, not just the proof of concept.
Work with engineering and product to put AI features in front of real users at scale, and stay accountable for how they behave once they're live.
Set and hold the engineering standard for production AI: appropriate safeguards such as content-safety guardrails, entitlement-based access control, evaluation, and model fallback, delivered as clean, tested, well-documented software with unit, integration, and end-to-end tests, containerized development, and CI/CD.
Use agentic software development tools such as Claude Code as a core part of the daily workflow, and set the team's practice for using them well.
Design rigorous evaluation frameworks for LLM and agentic systems — measuring accuracy, faithfulness, and robustness, and distinguishing acceptable model variability from factual error.
Critically evaluate scientific datasets and model outputs for quality, completeness, and scientific validity, applying domain understanding of chemistry, materials science, or the life sciences.
Establish trustworthy-AI practices that keep model outputs grounded in curated, provenanced scientific content rather than the model's own priors.
Independently synthesize technical findings and present strategic recommendations to senior executives and C-level stakeholders.
Influence organizational and product decisions through compelling, data-driven narratives and demonstrations.
Represent CAS's AI work externally through client engagements, conference presentations, and publications.
Set the technical direction for a team of data scientists and data engineers: the standards, patterns, and architecture for how agentic and LLM systems get built, spanning data and inference pipelines, model and prompt design, and deployment.
Own design and code review across that work, from model and prompt design through data pipelines and serving infrastructure, and raise the bar on quality, reliability, and reproducibility.
Mentor junior data scientists and AI/data engineers, and unblock them on difficult technical problems.
Set the team's standards for AI engineering, evaluation, and responsible deployment.
Lead and influence cross-functional initiatives as part of a shared-services model, without direct people-management responsibility.
Share knowledge across the team through demos, documentation, and internal learning sessions.