Luis Rosales spent two months engineering robust AI and data extraction features across the RoBorregos/home2 and RoBorregos/Home-Docs repositories. He refactored the Ollama-based deployment workflow, centralizing model management within Docker and streamlining entrypoint scripts for improved maintainability and cross-platform compatibility, particularly for macOS. In RoBorregos/home2, he made LLM-based data extraction the default, introduced fallback mechanisms, and enhanced NLP label categorization, leveraging Python and Shell scripting. For RoBorregos/Home-Docs, he standardized data extraction nodes and integrated FastWhisper prompts, improving user interaction. His work emphasized automation, reliability, and maintainable DevOps practices, resulting in faster, more reliable data processing workflows.