AI,M 9 21

Summary

The instructor explains how utility nodes help control and prepare data within a workflow. Unlike nodes that connect to external apps, utility nodes support workflow logic—for example, routing data down different paths based on conditions.

Practical tutorial: A profession-based learning-roadmap agent

The module builds on an earlier form-to-Airtable workflow and turns it into an AI agent that generates a customized learning roadmap.

1. Collect profession information

   • Add a Profession field in Airtable with options for Programmer, Designer, and Automation Expert.
   • Add a matching dropdown to the user form.
   • Map the new field into Airtable. When the value initially failed to save, the instructor refreshed the field mapping and corrected the form’s selection type to match Airtable’s multiple-choice field.

2. Route submissions with a Switch node

   • Add a Switch utility node and compare the submitted profession against each option using string equality.
   • Create one route per profession. The demonstration confirms that a submission follows the matching route rather than all routes.

3. Generate roadmaps with an AI Agent

   • Connect an AI Agent and a chat model. The instructor uses Google Gemini, selecting Gemini 2.5 in the demonstration.
   • The tutorial shows how to create an API key in Google AI Studio and emphasizes that API keys are secret and should not be shared.
   • Configure the agent to receive user details and act as a career mentor who creates a step-by-step learning roadmap. The instructor uses ChatGPT to draft the system prompt.

4. Save the AI response

   • Add a long-text Roadmap field in Airtable.
   • After the AI Agent runs, use Airtable’s Update Record action to pass the existing record ID and the agent’s output into the Roadmap field.
   • The instructor tests and revises the prompt when the agent does not initially recognize the learning topic. Further adjustments request Bangla output, a Salam greeting, and a more concise roadmap.

Key takeaways

• Use a Switch when multiple possible values need to trigger different workflow paths.
• Keep form fields, Airtable field types, and workflow mappings consistent.
• Test the workflow and refine prompts when the AI’s output misses an input or needs a different format.
• The finished example demonstrates a personalized roadmap generator, which the instructor presents as an introductory agent that can be expanded in later modules.

Main speaker and sources

• Main speaker: The course instructor from Ai Automation Mastery.
• Tools and services demonstrated or referenced: Airtable, Gemini via Google AI Studio, and ChatGPT for drafting prompt instructions.