The video tile is one of the most effective ways to reinforce ideas, increase engagement, and give your lessons a human touch. Whether you're guiding new hires through onboarding, demonstrating a new workflow, or breaking down a challenging concept, video can help learners absorb and retain information more effectively.
There are three ways to use the video tile in Mindsmith:
Upload existing content
Generate videos with AI
Build your own scenes from scratch, with AI assistance
Each gives you a different level of flexibility, depending on how fast you need to move or how tailored you want the content to be.
This guide walks through how each method works, what you can customize, and how to integrate video into your lessons for maximum impact.
Uploading or Embedding an Existing Video
If you already have a reliable video, such as a polished training segment, a recorded presentation, or a hosted clip, you can add it directly to your lesson.
To upload or embed a video:
In the Editor, click the Video button in the bottom toolbar.
Select Upload/Embed Video.
Choose from:
Upload a file from your computer
Paste a public video URL to embed
After adding the video, you can:
Trim unnecessary segments
Generate a transcript automatically
Upload a .VTT caption file for accessibility
Add a transcript manually
Choose a custom thumbnail image
Require learners to finish the video before continuing
Embedding is helpful when you’re combining established content with interactive or contextual elements. For example, you might pair a product demo video with questions that reinforce key takeaways.
Generating a Video with AI
When you don’t have video assets on hand, Mindsmith’s AI can create one for you from a simple prompt. This tool builds a full video with a virtual presenter, relevant imagery, and spoken narration—all based on the instructions you provide.
To generate a video:
Click the Video button
Select Blank Video
In the Create Video panel:
Enter a prompt or script (see below for guidance)
Choose the video length: Short (1–2 min), Medium (2–4 min), or Long (4–6 min)
Choose an image source: AI generated, stock, web-sourced, or none
Pick a virtual character to narrate the video
Click Generate Video
The system will generate your video in about 5–10 minutes. It will include narration, visuals, and a character delivering your script.
This method is useful when you want a quick way to personalize content, introduce new topics, or share thought leadership, without needing camera equipment or editing tools. It can also serve as a starting point you refine later.
Prompting Tips
If you’re not sure how to write an effective prompt for the AI video generator, try starting with a clear instructional goal and a few specifics. The more detail you give, the more tailored your video output will be. Here are a few examples:
“Create a 2-minute video introducing the fundamentals of inclusive leadership. Start with a definition, then explain why it matters in today’s workplace, and finish with a short story about a manager who modeled these principles.”
“Generate a short video to teach how to run an effective 1:1 meeting. Include three key dos and don’ts, and a quick scenario showing a manager adjusting their approach after feedback.”
“Build a 4-minute video explaining how to prioritize tasks using the Eisenhower Matrix. Use visuals to show each quadrant, and walk through a real-world example from project management.”
Want more guidance on prompting AI in the Mindsmith editor? Read our article on Getting the Most from AI Features.
Build Your Own Scenes from Scratch, with AI Assistance
If you want to design a video exactly the way you imagine it, scene by scene, Mindsmith gives you full creative control. This "Build From Scratch" option opens a blank canvas where you can shape the flow, visuals, narration, and structure in detail.
To start building your video from scratch:
In the Editor, click the Video button at the bottom of the screen.
In the Create Video modal, skip the prompt section and click Blank Video (Start from scratch) to create a blank slate video tile.
Your video will populate on-page. From there, select "Edit" (located on the tile) to open the video editor.
Once inside the editor, you’ll see a timeline interface where you can:
Add new scenes using the Add Scene button
Duplicate, reorder, or delete scenes using the timeline toolbar
Within each scene, drag and drop the following content elements:
Speaker: Choose a virtual character to deliver spoken narration
Text: Add written content like labels, titles, or onscreen guidance
Image: Upload your own visuals or search from AI, stock, or web sources
Video: Insert pre-recorded video files to layer in demonstration footage or B-roll
Customize each scene’s appearance:
Change background colors or upload a custom background image
Apply fade, slide, or zoom animations to content elements
Set timing and duration for each scene to control pacing
You can preview your video at any time using the play button, and when you're happy with a scene, simply move on to the next.
Using the AI Assistant When Building from Scratch
Earlier, we looked at how you can use AI to generate an entire video from a single prompt. But if you prefer more control over structure and tone, you can still use AI to support you while you build your scenes manually. The Assistant helps you fill in specific pieces—scripts, images, layouts—so you don’t have to start each scene from scratch.
When you're in the scene-based video editor, the AI Assistant appears as a prompt bar at the bottom of the screen. This is where you can type in instructions for what you’d like help with.
To use the Assistant:
Click the Assistant prompt bar at the bottom of the video editor.
[Insert screenshot: Assistant prompt bar location]Write a clear instruction. This can be for a new sequence of scenes, or a single element like speaker text or an image.
(Optional) Attach source links, adjust tone, or choose image types.
Submit the prompt. The results will auto-populate into your timeline, where you can continue editing.
Examples of helpful prompts:
"Create a 3-slide sequence introducing a new performance review framework. Slide one should define the framework, slide two should show an example of how it works in practice, and slide three should include a speaker summarizing key takeaways."
"Build a scene that introduces our company's new remote work policy. Include a professional background image, a speaker with a confident tone, and highlight two important changes in text."
"Write a closing scene for a lesson on giving constructive feedback. It should summarize the three-step feedback model, display the model visually, and feature a narrator wrapping up with encouragement to apply it in the next 1:1 meeting."
"Generate a slide that shows the difference between objectives and key results. Include one example for each, a side-by-side layout, and use clean, minimal visuals."
You can also fine-tune how the AI generates content by adjusting:
Sources: Add reference links for accurate or branded responses
Image preferences: Choose from AI generated, stock, web, or none
Speaker voice and language: Click the Voice option next to a speaker tile to change voice type or switch languages
Using the Assistant this way speeds up production without taking away control. You’re still the one building the scenes—but AI is helping you accelerate the repetitive or heavy-lift parts.
💡 Note: AI-generated videos will need to be rendered, which typically takes around 10 minutes. During this time, your video is processed and finalized.
If your video includes Speaker tiles, you won’t see any speech during preview mode. Once rendered, the full narration and visuals will appear.
Blending Video Methods in a Single Lesson
These three video tools aren’t mutually exclusive. You can use them together to shape a learning experience that fits your audience and content.
For example:
Use AI to create a welcome message that introduces a module
Embed a recorded video with subject matter expertise
Build a custom final video to summarize key points
Blending methods lets you align production effort with instructional goals. Use quick generation when you're short on time, and switch to manual controls when the details really matter.
Related Articles:
Tiles 101: Types of Lesson Content
Getting the Most from AI Features
Tiles to Know: Scenarios