Flock to Fedora 2026

Commercial AI assistants are impressive, but they run on distant corporate servers, harvest your data, and can change or disappear at any time. What if you could build your own AI chatbot that runs entirely on your machine, requires no GPU, and is yours to customize however you want?

In this hands-on workshop, you'll build a voice-enabled AI chatbot from scratch using open source tools on Fedora. No expensive hardware required, just a laptop running Fedora with 8GB of RAM.

What we'll build together:
* A local LLM using llama.cpp and a quantized model that runs on CPU
* Speech recognition using whisper.cpp so you can talk to your AI
* Text-to-speech output using espeak so your AI can talk back
* A customizable prompt to give your chatbot personality
* (Time permitting) RAG integration to give your chatbot knowledge from documentation

What you'll leave with:
* A working, talking AI chatbot running on your own hardware
* Understanding of the components that make up an AI assistant
* Ideas and tools to customize and extend your chatbot further

Who this is for:
Anyone comfortable with the Fedora command line who wants to understand how AI chatbots work under the hood, or who values privacy and wants an AI assistant they fully control. No prior AI/ML experience required.

Prerequisites:
* Fedora laptop with 8GB+ RAM and ~10GB free disk space
* Microphone and speakers/headphones

This workshop embraces the DIY spirit. Your chatbot is a starting point, not a finished product. Bring your curiosity and ideas for what you want your AI to do.