An autonomous content pipeline powering 12+ websites — generating 150+ articles/day with per-site tone control, model routing for cost-quality balance, cross-site linking engine, automated SEO optimization, and independent scheduling per site.
Built to scale content production across a network of 12+ niche websites without manual writing. The system uses LLM APIs with model routing — lightweight models for bulk content generation, premium models for quality-critical pieces — to produce 150+ articles/day, each tailored to the target site's audience and tone.
Each site runs as an independent scheduler process with configurable intervals. Content passes through quality gates before publishing. The system automatically handles WordPress publishing via REST API, generates thumbnails, manages internal cross-site linking, and generates comments for engagement. Daily health checks with per-bot Telegram alerting ensure continuous operation.
Each site runs as an independent scheduler process. A central orchestrator manages topic distribution, prevents content overlap, and enforces rate limits across all sites.
Content flows through a pipeline: topic selection → article generation (LLM API) → SEO optimization → thumbnail generation → WordPress publishing (REST API) → cross-linking update. State is persisted per-site with daily backups.