
XO — Minimal Conversational Intelligence Framework with Memory & Reinforcement Learning
XO is a lightweight, modular conversational AI framework designed for developers who want to build intelligent, context-aware, and self-improving chat systems. It combines short-term caching, long-term memory, and reinforcement-style learning to create natural, adaptive responses.
pip install xo-ai
git clone https://github.com/xo-aria/xo-ai.git
cd xo-ai
pip install .
from xo import AI
bot = AI("data.json")
response = bot.ask("hi")
print(response)
XO provides a built-in CLI for easy interaction:
from xo.cli import run_cli
if __name__ == "__main__":
run_cli()
And use cli in cmd:
xo run # Start an interactive chat
xo train # Merge long-term memory into database
xo reset # Clear short-term cache
xo help # Show command help
data.json automatically.{
"conversation": {
"hi": "hello there!",
"how are you": "Howdy! What can I do for you?"
},
"shop": {
"price": "The price varies by product. Which item interests you?"
}
}
XO was built to be easily modifiable:
You can modify these independently without breaking the framework.
$ xo run
XO is ready! Type 'exit' to quit.
You: hi
AI: hello there!
You: how are you
AI: Howdy! What can I do for you?
Created with ❤️ by XO Aria Designed for developers who believe AI should be both smart and simple.