Back to Services
Provisional course cover
Cover pending
Course

Generative AI for Biomedical Research

A practical course to learn how to use agents for research tasks, with a special focus on the biomedical field.

Format

Online or in person

Duration

6h

Audience

Biomedical research staff

Generative AIBiomedical researchLiterature reviewGDPR

Open notes

Access the course

An open preview of the course (only in spanish).

Read the notes

About this course

A practical, user-level course on generative AI applied to biomedical research. It focuses on understanding what these systems can and cannot do, organizing work with projects and sources, reviewing literature, writing better and delegating tasks to agents.

The specific tools may change between editions (and likely will), so here we prioritize more transferable training that lets you adapt to future changes.

Learning objectives

  1. 1.Understand the basic concepts behind language models, their reasoning levels and their limits in the biomedical context.
  2. 2.Tell apart model, chatbot, assistant, project, custom GPT, skill, connector, agent, orchestrator and automation.
  3. 3.Organize recurring documentation using projects and knowledge bases.
  4. 4.Build reusable skills for repetitive scientific work tasks.
  5. 5.Apply agents to literature search, review, synthesis and verification.
  6. 6.Use agents as assistants for drafting and revising scientific texts.
  7. 7.Get familiar with connectors, plugins, MCP, subagents and remote control for working with files, browsers and applications.
  8. 8.Apply ethical, GDPR and governance criteria when using AI with biomedical data.

Syllabus

Session 1 (3 h). Foundations, tools and knowledge organization

  • •Traditional vs. generative AI, pretraining, tokens, context window, biases and hallucinations.
  • •Model, chatbot, assistant, agent, harness and orchestrator: what each layer does.
  • •Reasoning levels, plan mode, usage windows and caching when switching models.
  • •Tool use, MCP, connectors and plugins; permissions, subagents and remote control.
  • •Projects, knowledge bases and persistent instructions with AGENTS.md and CLAUDE.md.

Session 2 (3 h). Reusable procedures and research work

  • •Custom GPTs, skills, reproducible prompting, automations and autoresearch.
  • •Scoping interview, protocol and two-pass literature search.
  • •Working with Zotero collections and PDF folders; reference extraction and verification.
  • •Drafting and reviewing in Word, Overleaf or local LaTeX, preserving voice and authorship.
  • •Patient data, GDPR, editorial policies and institutional governance.