Video Workflow

Create complete videos with AI-generated scripts, slide presentations, text-to-speech narration, and automated video composition.

Stages

Approval Gates

MP4

Output

TTS

Voice AI

Video creation wizard in the portal

Overview

The Video Workflow is the most complex workflow, combining multiple AI technologies to produce finished video content:

GPT-4 creates video scripts with scene breakdowns, narration text, and visual descriptions.

AI generates slide presentations with text, images, and visual elements for each scene.

OpenAI TTS converts narration text to natural-sounding voice audio for each slide.

FFmpeg combines slides and audio into final MP4 video with transitions and timing.

Gather video topic, style, and requirements

AI creates detailed video script with scenes

Human reviews and approves script

System creates slide tasks from script

AI generates visual content for each slide

Human reviews slides with feedback option

TTS creates voiceover for each slide

FFmpeg combines slides and audio

Final video available for download

Parameter	Type	Description
`video_title`	String	Title of the video
`topic`	Text	Main subject of the video
`duration_target`	Selection	"1-2 min", "3-5 min", "10+ min"
`video_style`	Selection	"explainer", "tutorial", "presentation", "promotional"
`voice_style`	Selection	"alloy", "echo", "fable", "onyx", "nova", "shimmer"
`slide_style`	Selection	"modern", "corporate", "creative", "minimal"

The video workflow requires FFmpeg installed on the server for video composition. Ensure ffmpeg is available in the system PATH.