Automatic Generation of Care Plans for Cancer Patients

This conference talk explores how Tenfifty and Regional Cancer Centers of Sweden are developing an AI solution to automate the generation of personalized care plans for cancer patients. Learn how their system uses Large Language Models (LLMs) in an agentic workflow to handle document structure, adapt existing content, generate new material, and provide necessary references—transforming healthcare professionals from writers to reviewers. Linus Johansson, a data scientist at Tenfifty with a master's degree in computer science from Chalmers University of Technology, details the technical implementation and challenges faced when applying generative AI to healthcare documentation. The presentation is particularly valuable for professionals working with generative AI and those interested in multi-step LLM workflows. Recorded at the 2025 GAIA Conference in Gothenburg, Sweden.