Skip to content

Instantly share code, notes, and snippets.

@donbr
Created January 20, 2026 22:43
Show Gist options
  • Select an option

  • Save donbr/021cf1f1121d5bc592028a5669dcc7e5 to your computer and use it in GitHub Desktop.

Select an option

Save donbr/021cf1f1121d5bc592028a5669dcc7e5 to your computer and use it in GitHub Desktop.
AI Makerspace Video Deliverable Guide (2026 Edition)

AI Makerspace Video Deliverable Guide (2026 Edition)

The Objective: Your video is an exercise in Developer Advocacy. It serves to prove two things:

  1. Functionality: You actually built the application.
  2. Grokking: You deeply understand the concepts and can teach them to a teammate.

1. The Setup: Anchor with a Diagram

Best for: Framing the narrative and helping those who find ad-libbing difficult.

Before diving into code or chat windows, start with a high-level architecture diagram (using tools like Excalidraw, Mermaid, or a simple flowchart).

  • Tell the Story: Use the diagram to visually explain the data flow—from the user query to the LLM to the final output.
  • Define Scope: This sets the stage for the viewer so they don't get lost when you switch between windows.

2. The Demo: The Hybrid Approach

Dr. Greg emphasizes blending the User Perspective with the Engineering Perspective.

Check out his Loom Video here: https://www.loom.com/share/70774f915c014da7b001e2e639f8fbb5

Part A: The Functionality (User View)

  • Show it running: Interact with the live application.
  • Example: Upload a document (e.g., Alice in Wonderland), ask a specific question ("What can we learn about rabbit holes?"), and show the system generating a response based on the retrieved context.

Part B: The Code (Engineering View)

  • Don't just scroll—explain: Pivot to your IDE to explain how the magic happened.
  • Key elements to highlight:
    • Imports: Briefly mention critical libraries (e.g., recursive character text splitters).
    • Prompts: Explain why you structured the system template the way you did.
    • Logic: Walk through the lifecycle—how files are ingested, how the state is managed on start, and how the chain executes.

3. The Personal Touch: "Lessons Learned"

Best for: Processing the material and differentiating your video.

Conclude or intersperse your demo with a personal reflection using the "3 Lessons" framework:

  1. What you learned: A specific technical concept that clicked for you during this build.
  2. What you haven't learned yet: Be honest about the "known unknowns"—areas where the code works, but you want to dig deeper (e.g., "I used this retriever, but next time I want to try a parent-document approach").
  3. Why it matters: Tying the new concepts back to what you already know.

4. Golden Rules

  • Keep it Short: This is a highlight reel, not a 2-hour lecture. A few minutes is sufficient.
  • Teach, Don't Read: Demonstrate that you understand the architecture well enough to explain it.
  • Full Stack Awareness: Briefly touch on the environment (Dockerfiles, deployment configs) to show you understand the full application lifecycle.

Pro-Tip: If you are looking for inspiration on tone, aim for a "Developer Advocate" style—enthusiastic, clear, and focused on the "why" behind the code.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment