The AI for Medical Affairs Course

• Key AI terminology, including Machine Learning, Neural Networks, Generative AI (GenAI) and Large Language Models (LLMs)
• Examples across R&D, clinical and commercial functions, with a focus on opportunities for Medical Affairs
• Demonstrations of reasoning models, AI agents and multimodal capabilities, including voice and vision

• How training data, supplied source material and access to additional information shape AI responses
• Managing the context window and understanding what your AI tool can see, remember and use

• Practical guidance on effective prompting, including defining the task, supplying source material and specifying the audience and desired output
• Working in small groups, participants use an available GenAI tool to test a task from their own work
• Reviewing responses, adding missing context and refining prompts to improve results

Findings and practical tips will be discussed in plenary.

• Prompt examples covering ad board outputs, literature review and synthesis, real-world evidence monitoring, and medical content review
• Applications in congress preparation, competitor intelligence and patient advocacy group (PAG) collaboration

• Using reasoning capabilities and reusable context to improve day-to-day outputs
• Working with files, projects and connected organisational information
• Practical examples, including reviewing changes to product information

• Screening and prioritising congress abstracts to identify relevant clinical data
• Supporting report generation and preparing resources for HCP engagement
• How faster turnaround creates more time for scientific interpretation and strategic decisions

Working in small groups, participants use AI to:

• Classify publications by study type
• Apply inclusion and exclusion criteria to identify relevant evidence
• Extract findings and draft a summary from supplied source material

The exercise explores consistency, traceability and the role of human review.

• Scientific accuracy, hallucinations, bias and referencing
• Data privacy, copyright, disclosure and regulatory considerations
• Prompt injection and the risks of relying on unchecked AI outputs
• Practical approaches to verification, accountability and responsible use as models and tools change

• Analysing content across channels to assess themes, relevance and relative impact
• Combining AI classification and scoring with human validation
• Using competitor benchmarking to inform investment and identify opportunities for differentiation

• Supporting discussion guide development and preparation
• Medical transcription, translation and live subtitles
• Linking insights to source quotations and producing outputs for internal use

• Personal AI assistants, developmental projects and enterprise solutions
• Assessing opportunities through efficiency, value and innovation
• Connecting AI implementation to strategic needs and practical workflows.

Participants use a structured planning framework to identify opportunities within their organisations. Working in small groups, they will:

• Identify business challenges and potential AI applications
• Use GenAI to explore and refine project ideas
• Prioritise opportunities by impact and feasibility, considering data requirements and strategic alignment

The selected ideas form the basis of the project development workshop on Day 2.

• Exploring the multimodal capabilities of Copilot, ChatGPT and Claude, alongside specialist tools such as Midjourney, Synthesia and ElevenLabs
• Applications in medical education, content creation and rapid prototyping
• Exploring medical slide deck creation and visual communication, including the need to check scientific accuracy

• Adapting expert-authored educational content into an interactive AI learning assistant
• Supporting medical review and testing responses before launch
• Using HCP questions and engagement data to identify educational needs

Building on Day 1, participants develop their prioritised project ideas. In small groups, they will:

• Clarify stakeholder needs, expected business impact and how AI will contribute
• Develop the proposed approach and identify how to test its value
• Use AI to support preparation and present their recommendations to the wider group

• How AI agents combine models, tools and repeated actions to work towards a defined goal
• Identifying suitable tasks and mapping workflows as standard operating procedures
• Defining data access, boundaries, human decisions and accountability
• Planning validation and monitoring for errors, model changes, performance drift and cost

A literature monitoring example illustrates how these principles apply in practice.

• Emerging AI capabilities, enterprise adoption and potential changes to Medical Affairs roles
• Assessing opportunities and limitations as AI develops
• Generative Engine Optimisation (GEO) and the visibility of medical information in AI-generated answers

Participants use AI-generated code and Google Colab to analyse a fictional dataset of Medical Science Liaison (MSL) reports. Working in small groups, they will:

• Explore reporting patterns and identify potential medical insights
• Investigate HCP questions, competitor activity and emerging themes
• Refine the analysis and check findings against the original reports
• Discuss which findings could inform Medical Affairs priorities and what further evidence is needed

• Applying the course learning to day-to-day work and future AI projects
• Identifying practical next steps for responsible AI adoption