Instructional Design Mentor
Overview
Title: Custom GPT: AI-Powered Instructional Design Mentor
Audience: Aspiring Instructional Designers, especially career changers from teaching, corporate training, or higher ed
Role: Instructional Designer, AI Prompt Engineer, Learning Experience Designer
Tools Used: ChatGPT (Custom GPT builder), Canva (branding), Miro (flow mapping), Google Docs (content design)
Categories: Scenario-Based Learning, Simulation, Microlearning
Problem & Solution
Many new instructional designers struggle to connect learning theory to real-world design work. While theories and models are widely taught, they are often presented in isolation, leaving learners unsure how to apply them when designing courses, building portfolios, or answering interview questions.
I designed a Custom GPT to act as a theory-to-practice bridge rather than a content delivery tool. The goal was not to teach theory exhaustively, but to help users reason through design decisions using theory as support.
The GPT functions as a guided learning companion, helping users interpret instructional design concepts in context and reflect on how their existing experience connects to professional ID practice.
Design Process
I began by identifying recurring pain points among aspiring instructional designers, particularly career changers. These included difficulty internalising theory, uncertainty around terminology, and lack of confidence applying concepts in practical situations.
From there, I:
- Mapped key theories and models to common ID tasks such as course design, needs analysis, and interview preparation
- Designed prompt logic to encourage explanation, questioning, and reflection rather than one-directional answers
- Defined clear boundaries for the GPT’s role so it supports thinking rather than replacing it
- Tested early versions with transitioning teachers and new designers to refine tone, pacing, and clarity
- Built flexible entry points so users could explore content by theory, scenario, or personal question
Reflection prompts were deliberately included to support metacognition and help users document how their thinking evolved over time.
Results & Reflections
Early testers reported increased confidence when discussing theory in interviews and applying concepts during project planning. The GPT proved most effective as a launch point rather than a destination, giving users language, structure, and examples that enabled deeper learning elsewhere.
This project reinforced my belief that theory-informed design does not need to feel academic or intimidating. It also deepened my understanding of how AI tools can be designed as constrained, intentional learning experiences that support adult learners without over-automating thinking.
© 2026 Flavia Bernardes