AI News Aggregator and Autopublisher
Role: Full-Stack Developer / Software Architect The technology stack: Laravel, PHP, API LLM (OpenAI/Gemini), WordPress REST API, Docker.
About the project: A fully autonomous system for collecting, processing and publishing news content. The architecture is built around the automation of content flow: from obtaining raw data to releasing a ready-made unique article with media materials without human intervention.
Key solutions and functionality:
-
Data aggregation: Configured to collect raw news from a pool of targeted sources.
-
AI rewriting: Integration with LLM for deep processing of source text, generation of SEO-optimized headers and preservation of semantic load at high uniqueness.
-
Graphics generation: Automatic creation of unique preview images (cards) under the context of each particular article through neural networks.
-
Publication pipeline: Integration with WordPress using custom console commands
news:auto-publish) for delayed or instant posting. -
Architecture and security: Strict network isolation (databases and internal services are limited to the Docker internal network, without binding to 0.0.0.0), strict filtering of CORS and the use of exclusively cryptographically persistent, generated accesses instead of default credits.