Frodex
An open-source platform for understanding LLM systems
Frodex is a practical learning environment where developers can explore how LLM-based systems actually behave — from tokens and prompts to RAG, tooling, and production patterns.
What is Frodex?
Frodex is a learning environment for people who build with AI every day — developers, technical PMs, data professionals, and founders who want to understand how LLMs actually behave inside real systems.
The content follows a structured path: from fundamentals like transformers and tokens to practices used in real products, such as advanced prompting, structured outputs, RAG, and production patterns.
Every lesson is designed to be practical: you understand the concepts and then immediately experiment in an integrated playground, with prompts and scenarios inspired by real application and system design problems.
The goal is to build practical intuition — not just learn concepts, but see how these tools behave once they become real systems.
Open source
Frodex is an open-source project. The codebase, lessons, and playground experiments are all available on GitHub. If you find a bug, have a suggestion, or want to contribute — pull requests and issues are welcome.
The platform is intentionally kept simple and modular so it’s easy to navigate and contribute to. Whether you want to fix a typo, improve an explanation, or propose a new module — there’s room for it.
About the creator
Diego Fedrizzi Petry Becker
Software engineer and technical leader focused on AI-powered systems
Based in Brazil.
I created this platform to help developers and technology professionals build practical intuition about how LLM-based systems really work — beyond the hype, with clear concepts, explicit trade-offs, and hands-on experimentation.
Let’s connect
These are usually the best places to connect and continue the conversation.
Quick contact
If you’d like to talk about applied AI, system architecture, or this project, WhatsApp is the fastest way to reach me.
WhatsApp