Building AI Agents shouldn’t feel harder than building the AI itself. Learn how AI Agents configuration can evolve into visual, maintainable AI workflows making intelligent behavior easier to design, understand, and scale in Drupal.
Basic knowledge of Drupal concepts such as modules, entities, and configuration. Familiarity with Drupal AI and AI concepts or workflow tools (e.g. n8n or Langflow) is helpful but not required.
As AI Agents grow more capable, configuration-based approaches quickly become difficult to reason about and maintain. Modern AI systems increasingly rely on workflows composed of prompts, tools, branching logic, and memory patterns popularised by platforms like n8n or Langflow.
This session explores how AI Agent creation in Drupal can be reimagined as AI Workflow rather than fragmented configuration screens. Using a real-world implementation powered by the the FlowDrop UI Agents module, we will demonstrate how AI Agents can be modeled as connected workflow steps, improving clarity, extensibility, and long-term maintainability.
Attendees will see how this approach aligns Drupal with modern AI orchestration patterns while preserving flexibility for developers and usability for site builders.
After this session, participants will be able to:
- Understand how AI workflow patterns apply to AI Agent design
- Identify limitations of traditional AI Agent configuration approaches
- Model Agent behavior using prompts, tools, branching, and state
- See how Drupal can support AI workflows similar to n8n or Langflow
- Apply workflow thinking to build clearer, more maintainable AI Agents
