Anyone can type a sentence into a video model and get eight seconds of something. The problem is that it usually looks like nobody decided anything — the subject drifts, the framing is accidental, the clip is pretty and says nothing.
That gap is not a tool problem. It is a directing problem. People who have never had to choose a shot size or an angle are suddenly being asked to describe one, and "make it cinematic" is the only vocabulary they have.
This programme teaches the visual language first, then the prompting. Participants learn composition, shot types and camera angles so they know what to ask for. They build a storyline that has a point. Only then do they write prompts, generate clips, and finish the audio and visuals. Idea, then image, then motion — in that order, every time.
It is deliberately not a course for creative departments only. A trainer building a module intro, an HR lead announcing a policy change, a founder who needs a product teaser by Friday — they all face the same eight-second problem.
Specific visual language beats vague "make it cinematic" prompts every single time.
This is a production journey, not a tour of tool tricks. Each module ends with a worksheet done on a real image, a real topic, or the participant's own project.
Participants analyse real photos and frames before writing a single prompt — eyes first, prompts second.
Composition, shot type, angle, storyline and prompt worksheets that carry the same project through all eight modules.
Prompts are run through current AI video tools in the room, then diagnosed when the output misses.
Full scene-by-scene teardowns, including a corporate training video built from storyline to finished clip.
Participants explain why they picked a shot or an angle. The reasoning is the skill; the vocabulary follows.
Each participant produces one complete AI-generated short video and takes the working files home.
At this programme's conclusion, participants should be able to:
Apply the core composition rules, and name them inside a prompt so the model actually delivers them.
Match shot size and camera angle to the question the audience is asking, instead of picking whatever looks dramatic.
Structure hook, problem, solution and call to action, then expand each act into visible scene beats a model can actually render.
Use the full component structure instead of a one-line description, so the same prompt produces the same quality twice.
Use locked subject descriptions and reference frames so three clips feel like one video, not three different worlds.
Fix audio clarity, colour drift, artefacts and pacing — the pass that turns an AI clip into a finished video.
Run a pre-production checklist covering message, audience, storyline, shot list, reference images and tool credits.
Each module pairs structured frameworks with reflective discussion and practical activity — so participants leave with usable tools, not just slides to forget.
Yes, for HRD Corp registered employers. As the training provider, we claim the course fee directly from HRD Corp. You submit the grant application before the training date and claim any other allowable costs, such as trainee allowances, yourself.
No. Trainers, HR, internal comms, sales and business owners get the most out of it, because they are the ones who need video but have no production team to hand it to.
None. The visual language is taught from scratch in the first three modules, which is exactly why the course runs two days instead of one.
Current-generation text-to-video and image tools, plus an audio tool for the finishing pass. These change every few months, so we confirm the exact stack before each run and teach the method rather than the buttons.
One finished short video, a filled-in workbook, a reusable eight-component prompt structure, and a pre-production checklist for the next project.
Two full days.
Speak to us about running this HRD Corp claimable programme in-house for your managers and team leads — or about scheduling a public cohort for your organisation.
Talk to Us