Architecting and generating an advanced UX Research course for a B2C scale up — leveraging the power of AI
The Client
Future Group
Future Group is a technology education and talent platform delivering high-quality coding and technology skills to working adults while providing businesses with a specialized tech-skilled workforce.
The Research Phase
Target Group
Young product professionals seeking career advancement
Young product professionals seeking career advancement within their organizations, professional development, or transition to new roles while building on their existing expertise.
Course Topic
Advanced UX (Methodologies & Practices)
Research findings were based on Google Search Volume analysis, Benchmarking, and a Survey distributed to the previously defined target audience.
Survey respondents showed interest in Advanced UX (methodologies & practices), soft skills, strategy and leadership, and data analytics.
Surprisingly, there was little interest in AI topics.
3 Deciding factors
Market Demand, Interest & Team Expertise
While the survey indicated strong interest in Soft Skills, Leadership, and Strategy courses, we needed to evaluate these findings against two crucial factors:
→ Market demand
→ Professional interest
→ Teaching team expertise
→ Our identity as a tech school
Pre-Production
→ Understanding the Knowns & Unknowns
We researched our direct competitors' courses, analyzing their content coverage and gaps, as well as current market demands for tools.
→ Due to budget and time constraints, we had to limit the scope of our research.
→ Nevertheless, our analysis proved valuable, with several of their topics serving as foundational elements in our curriculum development
Lessons Priorization
→ Using MoSKoW
During this process, we mapped not only the potential lesson content but also assessed how many coaches were available to teach each lesson at any given time.
Here we not only outlined what could be taught but how many coaches we had that could teach, at any time each one of the lessons.
Chapters & Modules Distribution
→ Finding Topic Affinities
After reaching agreement, we grouped all "Must Have" and "Should Have" topics into logical chapters within the established three modules, while setting aside the "Won't Have" and "Could Have" items.
We initially planned to include AI-related topics in all three modules, but ultimately this was only feasible in Modules 1 and 2.
The Content Generation Workflow Phase
We faced 3 Challenges
→ Limited AI Power At the time
→ SME Speed vs AI Gen Speed
→ Inconsistent Results from AI Models
1 Strategy Content Generation Workflow
→ Defining the flow of the whole Prompt Workflow→ Developing a system that broke down long texts into manageable chunks.
→ Crafting targeted prompts with clear guidelines, determining conversation breakpoints
→ Crated a self-review system for prompts to reduce our reliance on Subject Matter Expert (SME) input.
Prompt design
Summary, lesson, tone of voice, chapters, learning goals, reviews
We designed a comprehensive set of templates with detailed guidelines for future users.
The aim of this process was to achieve consistent results with our available tools.
We were using ChatGPT 4.0, which had more limited capabilities than the current models.
Send to SME for Review
Minimize SME involvement
The several times we sent our “ready to review” content to the SME, this one wrote several minor changes and topics to be added or removed.
The Auto-Reviewing prompts strategy accelerated our process greatly.
The Handover Phase
Create Playbook
Creating a Replicable Workflow Structure
The playbook provided a step-by-step guide for creating asynchronous course materials, with specific recommendations for best practices at each stage of content creation.
Playbook Goals
Establish a structured framework for creating high-quality course content using GPT-4
Provide step-by-step guidance for creating well-structured content across modules and chapters
Establish a clear, repeatable workflow that guides curriculum creators with detailed playbook guidelines and practical examples
Playbook Goals
Creating a Replicable Workflow Structure
→ Establish a structured framework for creating high-quality course content using GPT-4
→ Provide step-by-step guidance for creating well-structured content across modules and chapters
→ Establish a clear, repeatable workflow that guides curriculum creators with detailed playbook guidelines and practical examples
