n8n Workflow: Create Process Videos w/ OpenAI TTS + Metricool

Create Process Videos in 9:16 with OpenAI TTS + Metricool (n8n Workflow)

Turn an operator-submitted scene manifest and zipped images into a fully narrated 9:16 process video—complete with face-review validation—and optionally schedule it to Instagram, YouTube, and TikTok via the Metricool API.

What this workflow does

This n8n automation workflow generates process videos end-to-end, from asset intake to QA approval and publishing:

• Receives input from an authenticated n8n Form (or runs a manual synthetic demo): a scene-manifest JSON and a ZIP of images.
• Validates the manifest for required stages, correct chronological scene order, safe filenames, text length limits, and explicit reviewed face-box entries for every image.
• Uploads assets to a companion renderer service to generate masked preview images, then pauses for the operator to approve face coverage.
• Generates narration per scene using OpenAI Text-to-Speech, uploads each narration clip to the renderer, and starts the reviewed render.
• Polls render status until the renderer job completes or fails, then provides a review URL for operator QA and final approval.
• Optionally schedules publishing with Metricool: validates a UTC publish time plus caption/title, obtains the renderer’s public MP4 URL, schedules a three-network post in Metricool, and records scheduler ID/provider statuses (including a follow-up on a 15-minute schedule to fetch recorded scheduled posts).

Use cases

• Automation engineers need a repeatable way to produce process/explainer videos from structured scene data.
• SaaS operators want controlled face-review QA before exporting an MP4 for 9:16 vertical.
• Teams streamline content ops by using OpenAI TTS narration and Metricool multi-platform scheduling.

Technical details

• n8n nodes/logic: if, set, code, form, no op, wait
• Integrations/services: OpenAI Text-to-Speech, an external renderer service (render + review URL), Metricool API (Instagram/YouTube/TikTok scheduling)