Khabarkhan — AI Content Pipeline
The AI pipeline running this very site: it reads RSS feeds, writes original bilingual articles and generates its own cover art.
- Role
- Creator
- Year
- 2026
- Status
- live
Khabarkhan is the automation behind this site's blog. A scheduled job pulls from curated RSS sources across AI, frontend and backend, picks the freshest unprocessed item per topic, and asks an LLM to write an *original* article about it — not a translation or a summary — in both English and Persian.
It then generates a matching cover image with an image model, uploads it to object storage, and files the post as a draft for review. Nothing publishes itself without me approving it.
The whole thing is sized to run inside a 60-second serverless limit, so it does exactly one unit of work per invocation and is driven by an external cron.
Highlights
- RSS → LLM → original bilingual article (not a translation)