Building My Own AI Motion Graphics System

Internal Tooling — AI + Motion

Building My Own AI Motion Graphics System

← All work
001 / INTERNAL SYSTEM

I didn't want another AI tool that only gives ideas. I wanted a system that could actually build motion.

CLAUDECUSTOM MCPLOCAL RENDERINGCMD WORKFLOW

THE SYSTEM / 01

Prompt in.
Motion out.

A custom local pipeline connecting Claude, MCP and a purpose-built rendering workflow.
FARHAN MOTION / LOCAL
SCENES01_HOOK02_BUILD03_REVEAL04_OUTRO
01scene.define({ duration: 10 })
02mcp.connect(claude)
03timeline.map(visuals, audio)
04render.queue('vertical-4k')
Rendering scene_03…
ZOAVE
03
MOTIONENGINEGENERATED / LOCAL
01 / COMMAND LAYER

Claude became
the interface.

Plain language enters the system. MCP turns the direction into structured instructions.

CLAUDE / DIRECTION

Create a cinematic 10s product scene with a clean reveal and controlled camera motion.

SEND TO MCP
MCPTRANSLATE
scene.jsoncamera.jsontiming.json
02 / LOCAL ENGINE

CMD became
the production floor.

The generated scene files feed directly into the local project, where the visual render is executed.

LOCAL / CMD
> zoave run scene_01.json> loading assets... 100%> building timeline... 100%> rendering frames... 100%> export visual-pass.mov ✓
SCENESASSETSRENDERVISUAL PASS
03 / AUDIO LAYER

Voice and sound
join automatically.

The system can create the voice track, find free-to-use sound effects, and build a timed audio layer around the motion.

AUDIO / AUTO MIX
VOICEAUTO-GENERATEDSYNCED
01Analyzing scene rhythm✓02Generating voice track✓03Matching free-to-use SFX✓04Balancing mix to picture✓
VOICEAI NARRATIONSFXFREE-TO-USE LIBRARYMIXAUTO SYNC
04 / FINAL OUTPUT

One system.
Complete motion.

Direction, animation, rendering, voice, sound design and export become one continuous production flow.

MOTION + AUDIO / TIMELINE
ZOAVE
FINALMOTION
OUTPUT
VOICE + SFX / READY

WHY THIS IS A PAID SYSTEM

You are not paying for a prompt.

You are paying for the working production system that came after the prompts.

01Time

Architecture, scene logic, integrations and production workflow took real build time.

02Iterations

Many approaches were tested, rejected and rebuilt before the workflow became reliable.

03Token cost

Development consumed substantial AI context, generations and repeated debugging cycles.

04Final system

The product packages the proven workflow so you can start from the result instead of repeating the experimentation.

THE VALUE

Skip the months of trial and error. Use the system we already spent the time, iterations and tokens building.