AI Content Creation
Fifteen days from a written brief to one finished cinematic short — specified, generated and mastered under direction.
- Starts
- 15 September 2026
- Venue
- Coimbatore
- Duration
- 15 working days
- Contact hours
- 30 hours, 2 per day
- Batch size
- 15 students
- Fees
- From ₹4,000
Who it is for
No prior editing, photography or film experience is assumed. You need working computer literacy, the ability to write and read a script in the language of the batch, and your own configured device.
Apply for this course
What you will be able to do
- Convert a content brief into a locked script and a shot list carrying camera, lighting, duration and continuity parameters for every shot.
- Construct controlled prompts using role, context, constraint and output-schema specification, and maintain a versioned prompt library.
- Generate character-consistent and environment-consistent image assets, including rendered in-image text, at production resolution.
- Generate video from text, from a single keyframe, from first and last keyframes and from supplied reference assets, choosing the model tier to suit the shot and the credit budget.
- Diagnose a failed generation and correct it by isolating prompt, reference or model as the cause.
- Produce and synchronise narration, sound design and music to a defined loudness target.
- Assemble, grade, caption and master a finished film to platform delivery specification, and derive a vertical cut by reframing rather than cropping.
The programme
Specification, prompt architecture and asset design
Days 1–7 · 14 contact hoursNo video is generated in this module. Students who begin generating on day one accumulate material they cannot assemble.
- Generative video production systems, the six-stage pipeline and credit economics
- Concept development and audience specification by problem, desire, objection and trigger
- Prompt engineering I: role, task, context, constraints and output schema; versioning and library discipline
- Script, hook and narrative structure for a 45 to 60 second film, timed against a beat sheet
- Shot design: size, angle, lens, movement, lighting vocabulary, palette and continuity rules
- Prompt engineering II: image specification, character sheets and rendered in-image text
- Consistency, single-variable iteration and asset lock; first and last keyframes for every shot
Generation, post-production and mastering
Days 8–15 · 16 contact hoursDirected generation: every clip is produced against a written specification and measured against it.
- Workspace and directed generation; model tier selection and the test-low then promote method of credit control
- Keyframe and reference control; clip extension without continuity break; building a sequence rather than isolated clips
- Shot recovery: reframing, re-timing, substitution, upscaling, and when to abandon a shot and rewrite the edit
- Voice, language and performance; pronunciation, pause, emphasis and pacing, with Indian language and accent handling
- Assembly and edit craft: voice-first assembly, J and L cuts, cutting on action, match cuts
- Pacing, titles and captions; typographic hierarchy, safe areas, kinetic titles and brand lock-up
- Sound design, colour and finishing; music licence status, ducking, loudness targets and grade
- Mastering and delivery: 16:9 master, 9:16 derivative, thumbnail set, archive and assessed review
Assessment, out of 100
- Prompt engineering20
- Concept and script15
- Visual asset quality and consistency15
- Generation control15
- Edit craft15
- Voice and sound design10
- Delivery compliance10
Prompt engineering and generation control together carry 35 marks, more than the finished edit, because those are the transferable competencies.
What you leave with
- Creative brief and selected concept with written justification
- Locked script, beat sheet and parameterised shot list
- Versioned prompt library, image and video prompt sets
- Character sheet, environment plates and locked keyframe set
- Generation and credit log, including rejected shots and the reason
- Narration, music and sound effect assets, and the editable project file
- Final 16:9 master of 45 to 60 seconds, plus the 9:16 social cut
- Thumbnail and cover frame set, and the archived project folder
Production stack
- Concept, script and prompt engineeringChatGPT
- Image generationNano Banana 2
- High-fidelity plates and typographyNano Banana Pro
- Video generationGoogle Flow, running Veo 3.1
- Alternate generationRunway, Kling, Sora
- VoiceElevenLabs
- Edit, captions and finishCapCut Pro
- Optional finishingDaVinci Resolve
Model versions and plan entitlements in this field change on a scale of weeks. The course is built around the pipeline stages, not the application names, and an equivalent is substituted where a tool has moved.
What you need to bring
- Your own workstation, 16 GB RAM minimum, with 100 GB free storage
- Closed-back headphones — mixing assessed on laptop speakers is not accepted
- All accounts provisioned, logged in and credit-verified before day one
- 45 to 60 minutes of directed self-study each day, roughly 13 hours across the programme





