01
Manual every day
Opening the same upload dialogs, pasting titles and descriptions, checking boxes, waiting. Twenty minutes gone — every single day.
Automation · Local · Free
Bulk YouTube
upload pipeline.
I was wasting 20 minutes every day manually uploading 6 videos. Built a fully automated engine that runs on my laptop — $0, 7 nodes, zero cloud.
Context
The friction. Then the fix.
The same workflow, redesigned with the visual language of the uploaded Be icon: luminous blue glass, polished highlights, depth, and precise geometry.
02
Local automation
A self-hosted n8n workflow that reads files from disk, injects metadata, respects API limits, and notifies when everything is live. No SaaS. No cloud bill.
Process
How it works.
01
Local n8n setup
Pasted my PC specs into Claude and asked it to set up n8n for free. Ten minutes later the engine was installed and running on my machine.
02
Logic with Gemini
Hit Claude’s limit mid-build, pivoted to Gemini to design the actual workflow logic and node configuration.
03
API-aware spacing
Reads MP4s from the hard drive → splits into 6 upload jobs → injects titles & descriptions → sets “Not Made for Kids” → waits 2 minutes between each upload so YouTube never blocks.
04
Telegram notification
Final node is a Telegram bot. When the sixth video goes live, the phone buzzes. The process runs without needing attention.
Architecture
The workflow.
LOCAL AUTOMATION / N8N
Schedule → Read files → JavaScript → Loop → YouTube Upload → Wait → Telegram
Stack
Tools used.
n8n (local)
Claude
Gemini
YouTube API
Telegram Bot
Own machine