Generic course creator simulation • current local flow
This page shows how an instructor should move through CourseForge for any course topic: login, describe the course, review AI outputs, work through each AI-generated session, review individual slides inside that session, request changes if needed, approve each stage, and end with a voice/video production plan.
The instructor does not manually edit everything. They provide direction, review what AI creates for their specific topic, then approve or request revision.
The instructor starts from the login screen. For local testing, use the seeded test account.
After login, the instructor gives the AI enough direction to create useful course assets. The current MVP uses structured fields; the intended next UX is a chat-style intake that fills these fields in the background.
The app turns rough instructor notes into course alignment cards: promise, audience, source material, lecture ideas, and recording choices.
Review loop: instructor can type a note like “make it more workshop-like” and ask the AI to revise before moving forward.
Instructor action: Align goals + generate lecture planThe instructor reviews lecture goals and the structure before slides are created.
Aligned course goal Learners can run five useful customer interviews and turn insights into product decisions Lecture 1: Set the transformation Goal: Align founders on the outcome. Plan: Open with learner pain and define success. Lecture 2: Teach the core framework Goal: Help founders understand the repeatable customer discovery sequence. Plan: Break interviews into a simple framework. Lecture 3: Apply with examples Goal: Show a practical walkthrough using instructor material. Lecture 4: Handle mistakes Goal: Show traps, edge cases, and self-correction cues. Lecture 5: Finish with action Goal: End with a checklist learners can execute after 12 minutes.
The deck is organized by session. Every session now uses the backend AI slide-generation endpoint, so the platform is not limited to Session 1 or to a fixed course topic.
Review loop: instructor can request “simplify slide S2.2” or “make Session 3 more practical”, then approve slides one by one. The AI regenerates content from the instructor’s topic, not from a hardcoded test course.
Instructor action: approve session slides + create scriptThe instructor reviews the voiceover script. Every approved slide is sent to the backend script-generation endpoint, so the script covers the full course, not only the first session.
Slide S1.1: Why customer interviews matter Before you write another feature request into your roadmap, pause and ask whether you have heard the customer's real problem in their own words... Slide S1.2: The five-question interview flow Start with context, then ask about the last time the problem happened. Avoid pitching your product too early... Slide S2.1: Turn notes into decisions After each interview, separate facts, quotes, assumptions, and product decisions. This keeps the team from overreacting to one loud opinion... Slides continue across every approved session Each approved slide becomes a separate narration block for voiceover.
Review loop: instructor can ask for warmer tone, less jargon, shorter script, or more examples.
Instructor action: Approve script + prepare voice/recordingOnce approved, the app knows what assets need to be created and where they should live. Backend production endpoints for local TTS and final browser-recorded video are still the next implementation slice.