I make quality awareness training videos as scenes: a small cast of characters on a factory floor or in a workplace, speaking to each other and doing the work, kept consistent across a whole series. The scripts come from you, and the characters, voices and locations are made with AI image, video and voice models. It's one of the kinds of training video I make as part of AI video production.
Quality awareness and ISO 9001 awareness training
Quality awareness training is usually a set of short videos, each showing one idea in a situation staff will recognise: checking a part against the drawing, stopping to raise a non-conformance, handing over a job with the paperwork complete. The same characters appear across the set, so each new video picks up where staff already know the people and the place.
Workplace behaviour training
Workplace behaviour and code of conduct training uses the same method: characters in a scene, a situation, and how it plays out. Scenes like these are where a cast that stays the same matters most, because the viewer is following the same people from one video to the next.
How a scene with several characters gets made
Each character is set up once, from reference images of their face and clothing and a sample of their voice. Every shot they appear in is generated from those same references, then checked against them before it goes into the edit.
Two characters speaking in the same shot is the hardest case for current video models, and one speaker per shot is far more reliable. So a conversation is built the way a filmed one is cut: a sequence of shots, each with the character who is speaking, and the voice lines cut to land on the shot they belong to. Some scripts written as one continuous two-person exchange get cut a little differently once they're on screen, and that's agreed at the shot-list stage.
What I need from you
- The scripts, or the points each video has to cover so I can write them.
- The characters, if they already exist as images, or a description of who they are so I can design them.
- The voices, if they already exist, or a description of how each character should sound.
- Photos of the workplace, so the floor, the uniforms and the equipment match.
Tell me about the training.
Send the scripts or the points the videos have to cover, and I'll come back with a quote against scope.
Ask for a quote →If the scripts are ready, email them straight to advisory@danielbilsborough.com and I'll quote from those.
Questions about quality training videos
Can you use characters and voices we have already made?
Yes. Character images become the reference for every shot they appear in, and voice samples become the reference for how each character sounds. If the characters or voices were made in another tool, that tool's terms decide whether they can be used outside it, so that is worth checking first.
How consistent are the characters' faces from one video to the next?
Each character is generated from the same reference images in every shot, and every shot is checked against those references before it goes into the edit. A shot where a face has drifted is made again.
Can quality training be made as a series of short videos?
Yes. The cast, the workplace and the voices are kept from the first video, so each later video in the series starts from the same characters and places.
Does an AI training video count as ISO 9001 training evidence?
The video is training content. Whether watching it counts as evidence of training is decided by your quality management system and your auditor, usually alongside a record of who watched it and a check of what they understood.
HeyGen supports up to three presenters sharing a scene, built around people speaking to camera. What HeyGen does and where it stops covers it. A different kind of training video, where the machine is the subject, is on the equipment maintenance training videos page, built the same way as a nine-minute maintenance film I'm producing now.