How We Create Content
Transparency about our research and AI-assisted publishing pipeline.
Content on this site is produced through a research-to-publication pipeline that combines human judgment with AI tooling. This page explains each step.
1Research
The author uses Gemini Deep Research and NotebookLM to conduct in-depth topic research. Resulting documents are stored and uploaded to Google Drive.
2Knowledge Base
Research documents are ingested into a Weaviate vector database (semantic search index) using Gemini embeddings. Combined with official vendor documentation servers, this forms a unified knowledge base the system queries when drafting new articles.
3Content Drafting
When an article idea is selected, the system queries a unified documentation service that searches official vendor documentation from Microsoft Azure, Google Cloud, AWS, Cloudflare, and OpenAI via their Model Context Protocol (MCP) servers alongside the Weaviate knowledge base of ingested research documents. The results are enriched with live Tavily web search results, and the combined context is passed to Gemini (Flash model) to draft the article. A second Gemini Pro pass acts as editor to polish tone and consistency.
4Human Review
Every draft is reviewed by the author before it is approved for publication. No article is published without editorial sign-off.
Human oversight is essential for deciding what to publish and for applying context that models alone cannot provide.
5Images
Article images (hero and supporting) are generated by Google's Imagen 3 model via Vertex AI, based on prompts derived from the article content.
6Translation
The approved English article is translated into French, German, Spanish, Italian, and Dutch using Gemini Flash. The original English version is the source of truth; translations are provided for accessibility.
Tools & Technologies
The pipeline uses: Gemini Deep Research and NotebookLM for research; Google Drive for document storage and ingestion; Microsoft Learn MCP, Google Developer Knowledge MCP, AWS Knowledge MCP, Cloudflare Docs MCP, and OpenAI Docs MCP for official vendor documentation; Weaviate for vector search over ingested documents; Tavily for live web research; Gemini Flash and Gemini Pro for drafting and editing; Imagen 3 (Vertex AI) for images; Gemini Flash for translation.
Questions
For more about who runs this site and how to get in touch, see the About page.
About