Agentic AI Workflows: Building Multi-Step Workflows

AI is moving beyond answering questions and into taking action. This three-hour session introduces learners to agentic AI: systems that can plan, sequence, and carry out multi-step tasks with minimal manual intervention. Using accessible no-code or low-code platforms such as Make.com or n8n, learners will explore how to design workflows that connect AI capabilities with the tools and data sources already in use at Stanford. This session is ideal for staff who want to reduce repetitive work and begin experimenting with automation in a hands-on, practical setting.

Learners will have the opportunity to:
1. Explore how agentic AI differs from traditional chatbot or single-step AI interactions
2. Work with a no-code or low-code workflow platform to design and test a multi-step automated process
3. Practice connecting AI actions to triggers, conditions, and outputs across different applications
4. Experiment with common workflow patterns such as data routing, summarization pipelines, and notification automation
5. Identify repeatable tasks in your own work that may be good candidates for agentic automation

Topics:
- What agentic AI is and how it differs from prompt-and-response AI
- Overview of workflow automation platforms: Make.com, n8n, and related tools
- Core concepts: triggers, actions, conditions, loops, and branching logic
- Connecting AI models to external apps, forms, email, and data sources
- Building and testing a multi-step workflow during the session
- Common use cases: document processing, routing and notifications, research pipelines
- Guardrails and considerations: when to automate and when to keep humans in the loop