The Workflow Edit | Practical AI Tool Tutorial
How to Use HeyGen Without Generic AI Instructions
Stop hiring presenters and setting up cameras. Create a short training video with an AI avatar that reads your script with natural movement.
Dear Suzannah
Dear Suzannah, I need to create a short training video for new hires but I do not want to be on camera and I cannot afford to hire a presenter. The last time I tried a talking-head video, it looked like a stiff mannequin reading a script. Is there a way to make a presentable video without filming myself?
Pick a stock avatar and paste your script.
HeyGen lets you choose from stock AI avatars, paste your script, and generate a video where the avatar speaks your words with synced lip movement. You get a presentable training video without a camera, lighting, or a second take.
The real use case
You need a short training or onboarding video but you do not want to film yourself and hiring a presenter is not in the budget. You pick a stock avatar in HeyGen, paste your training script, and the tool generates a video where the avatar speaks your words with natural lip sync. You review the result, fix any pronunciation issues, and export a video your new hires can watch on day one.
The tool-specific prompt to use
Paste this prompt into HeyGen’s script input along with your training content. Replace the bracketed items with your real video details before generating.
Create a [length, e.g., 90-second] training video for [audience, e.g., new hires, field staff, clients]. Avatar: select a [professional, friendly, neutral] stock avatar that fits a [corporate, casual, educational] setting. Avatar attire: [business casual, uniform, smart casual]. Voice: [calm and clear, energetic and warm]. Language: [English]. Script topic: [what the video covers in one sentence, e.g., how to log expense reports]. Paste the full script below and assign it to the avatar as a single continuous take. Add on-screen text for these key terms when they appear: [list three to five terms, e.g., receipt, approval, deadline]. Background: [plain office, solid color, branded gradient]. Do not add background music unless the video is over 60 seconds — if so, use a soft, low-volume track. Add a title card at the start showing . Add an end card showing [next step, helpdesk link, or contact]. Keep the avatar centered and at eye level for the full video. If the avatar mispronounces a [product name, acronym, or proper noun], note it and regenerate that line. Script: [paste your full training script here]. After generating, review the full video once and flag any line where the lip sync looks off.
Prompt length: 197 words.
Make the result less generic
- Choose an avatar whose style matches your workplace, not the default option.
- List the exact key terms that need on-screen text so they appear at the right moment.
- Specify a plain or branded background instead of a busy scene that distracts from the speaker.
- Note any product names or acronyms that might be mispronounced so you know what to check.
- Keep the script under two minutes so the avatar movement stays natural throughout.
Quick human check
- Does the avatar look natural and not stiff or mannequin-like?
- Is the lip sync accurate, especially on longer sentences?
- Are all product names and acronyms pronounced correctly?
- Do the on-screen text overlays appear at the right moments without blocking the avatar?
- Would a new hire understand the training after watching this once?
