CONNECT THE REPETITIVE WORK
Internal tool developmentInternal AI tools
Content Automation Workflows
Match each section with the right video scene.
A production tool that matches article headings to video transcripts, extracts relevant frames, and inserts them into the article.
HOW IT COMES TOGETHER
The core workflow- 01
Article and transcript
- 02
AI scene matching
- 03
Frames and uploads
- 04
Images in the article
01 / THE QUESTION
The problem we set out to solve.
When creating an article from a video, finding, capturing, and uploading a suitable scene for each section is simple but time-consuming work.
02 / OUR APPROACH
Our approach.
Article headings and a video transcript are used to locate related moments. Frame extraction, storage uploads, and HTML insertion then complete the repetitive steps.
- 01
Read article structure
HTML section headings are extracted to identify the topic of each passage.
- 02
Find relevant timestamps
An LLM compares the meaning of a heading with the transcript to choose a related video segment.
- 03
Extract and upload frames
A frame from the selected moment is captured and uploaded to storage.
- 04
Assemble the article
Images are inserted beneath their headings and connected to an n8n production workflow.
03 / INTELLIGENCE AT WORK
Where AI does the work.
The LLM matches the meaning of text and video transcripts. Automation tools perform the downloads, captures, uploads, and HTML editing.
04 / WHERE WE ARE
Where the project stands.
An internal toolkit combining an n8n environment with Python production scripts. It supports content production rather than operating as an independent consumer service.
