Problem
Editorial team was hand-translating Word documents into web pages — slow, error-prone, and a constant bottleneck before any campaign launch.
Approach
Built a parser that ingested the Word source, mapped its structure to a strict template, and emitted production-ready HTML with consistent metadata and styling.
Outcome
77 pages generated in a fraction of the previous time. The editorial team kept ownership of the content; the pipeline owned the translation layer.
Impact
- 77Pages auto-generated
- −90%Publishing-time reduction
Tech stack
Python
Docx parsing
HTML/CSS generation