Here will be centralized all the old and current posts and thoughts, with tips and some solutions that I have implemented throughout my career.
You don't need a $1,000 GPU to run AI locally. With the right tools and small models, you can have a coding assistant running on your college laptop.
The self-attention mechanism that powers modern AI has a dirty secret: most of its calculations are redundant. A simple mathematical insight turned this weakness into a 10x speedup.
The 'Attention Is All You Need' paper didn't just improve AI—it threw out the rulebook. Here are the radical ideas that quietly reshaped everything we know about artificial intelligence.
The real problem isn't context window size - it's how the model interacts with information. Recursive Language Models propose a paradigm shift that could redefine what LLMs can actually do.
What they don't tell you in tutorials about building production-ready RAG systems. Lessons learned from real deployments that separate working systems from failed experiments.
Discover how to transform your development workflow using Claude Code's advanced features like custom skills, parallel agents, and superpowers that turn vibe coding into a productivity machine.
A comprehensive analysis of the architectural decisions behind migrating from a Java-based processor to a modern Elixir Phoenix application
Complete guide on how to set up and use Dokploy to simplify deployments on VPS or homelab - the open source alternative to Vercel and Netlify for your own infrastructure.
Discover how to use Claude AI as a DevOps assistant for personal projects, automating infrastructure and solving complex problems efficiently and cost-effectively.
Learn how to create a robust, real-time chat interface using Phoenix LiveView integrated with OpenAI's streaming API, featuring comprehensive error handling and a polished user experience.