Real AI Engineering projects built during ITAnna sessions — study the code, fork it, build on it.
Can a neural network discover a hidden rule that humans can't easily guess? 🧠 We build "PatternMachine" — a minimal neural network (just 2 parameters) to demystify how models learn mathematical rules from data, watching weights converge to the exact ground truth formula in real-time.
View on GitHub →Hands-on prompt engineering examples — zero-shot, few-shot, chain-of-thought, and structured outputs.
Coming SoonBuild Retrieval-Augmented Generation (RAG) systems from scratch, implement vector databases, and perform semantic search.
Coming SoonBuild an orchestration system using LangGraph to visualize agent decision paths, control loops, and multi-step workflows.
Coming Soon