Senior Data Scientist – Build the AI That Powers Life-Saving Medicine Production

Data, Analytics & AI, Sweden Operations - Södertälje, Sweden

Do you want your models, agents and AI systems to run in the factory that makes medicines for millions of patients?

At AstraZeneca, we follow the science and put patients first. We are building the next generation of AI in pharmaceutical manufacturing and supply, and we are looking for a Senior Data Scientist who can help us take it from promising prototype to trusted production reality.

About the role

Södertälje is one of the largest sites in AstraZeneca’s global network and a recognised World Economic Forum Global Lighthouse for digital manufacturing. Sweden Operations is now moving beyond individual predictive models towards connected, AI-enabled operations: decision support that senses what is happening across the value chain, reasons about it and recommends or takes action, with people in control.

As Senior Data Scientist in the Data, Analytics & AI team, part of Strategy & Performance, you will be a technical leader in that shift. You will work with colleagues across Inbound, Manufacturing, Quality, Outbound, Maintenance, Supply and Production Planning to turn complex operational problems into AI solutions that are reliable, explainable and compliant, and that people actually use.

What you’ll do

• Design and build AI solutions end to end. Combine machine learning, optimisation, simulation and generative AI to solve real operational problems, from forecasting and anomaly detection to document intelligence and process optimisation.

• Develop LLM and agentic AI applications. Build retrieval-augmented and agent-based systems that work with our data, tools and procedures, with rigorous evaluation, guardrails and human-in-the-loop design.

• Frame the problem before the model. Translate unstructured business challenges into the right analytical approach, including when the answer is a simple model rather than a complex one.

• Take solutions into production. Own the path from prototype to a monitored, maintained service using modern MLOps/LLMOps practices on our cloud and data platforms.

• Make AI trustworthy. Build quality, explainability, validation and documentation into the work so that it meets GxP, data integrity and regulatory compliance.

• Influence. Communicate results, uncertainty and limitations clearly to stakeholders from the shop floor to senior leadership and help them decide where AI will and will not add value.

• Keep the team ahead. Follow research and tooling developments, test what is useful, and bring it into our ways of working.

Who you are

You are curious, rigorous and pragmatic. You like hard, ambiguous problems, but you care more about measurable impact than about novelty.

Essential

• Master’s degree (or equivalent experience) in mathematics, statistics, computer science, engineering, physics, operations research or a related quantitative field.

• Several years of experience delivering data science solutions into production, with a track record of business or scientific impact.

• Strong Python and SQL skills, with solid software engineering practices, including Git-based version control, testing, code review, CI/CD, Docker, and Kubernetes.

• Strong data engineering skills, including building and maintaining reliable data pipelines, working with complex data sources, and ensuring data quality for analytics and AI applications.

• Deep knowledge of machine learning and statistics, including supervised and unsupervised learning, with experience in some of time series, forecasting or deep learning, and sound judgement on model choice, validation and uncertainty.

• Hands-on experience building LLM-powered applications, including RAG, tool calling, and agentic workflows and orchestration, with a disciplined approach to evaluation, regression testing, and monitoring.

• Experience deploying and operating models and AI services in a cloud environment, including CI/CD, containerisation, monitoring, and pipeline orchestration.

• Excellent communication and stakeholder skills, and the ability to explain technical work in business language.

• Experience using AI-assisted development tools such as GitHub Copilot, Claude Code, or similar, following best practices for code review, testing, and security.

• Fluent English, written and spoken.

Desirable

• Optimisation, simulation, digital twin or process modelling experience; computer vision or multimodal AI.

• Experience in manufacturing, supply chain, life sciences or another regulated environment (GxP, validation, data integrity).

• Knowledge of responsible AI, model governance and risk management, including the EU AI Act.

• Experience with Snowflake or similar data platforms, orchestration tools, and agent frameworks.

• Experience coaching or leading other data scientists.

• Swedish language skills.

• Product management skills, for example shaping value cases, prioritising a backlog and working with business owners to drive adoption.

How we work

We follow the science: we test our assumptions, measure what matters and share what we learn. We put patients first: reliable supply of quality medicines is the reason every model we build exists. We play to win and do the right thing: we set ambitious goals, and we never trade away quality, safety or integrity to reach them. We are entrepreneurial: we try new ideas quickly and learn from them, and we are happy to retire what does not work.

Why AstraZeneca?

You will work on problems with real consequences, with access to large-scale, high-quality operational data and a company that has made AI a strategic priority across the whole business. You will join an international, collaborative environment with strong opportunities to learn, to publish or present your work and to grow your career. Diversity and inclusion are at our core, so that everyone can bring their whole self to work.

What’s next?

Apply with your CV and a short cover letter as one combined document no later than November 19, 2026. In your letter, tell us about one AI or data science solution you took into real use: the problem, your approach, how you evaluated it, and what you would do differently. If you have public repositories or a portfolio, please include a link.

Join us, and help turn data and AI into better, faster and more reliable medicines for patients. We look forward to hearing from you.

Date Posted

Closing Date