The pieces of your life.
Messages, mail, photos, voice notes, and moments. Choose the sources that tell your story.
A PERSONAL AI, IN THE MAKING
The way you think. The words you choose.
The little things that make you, you.
Meet the AI we're building around all of it.
Your own model. Your own context. Wherever you go.
Every new assistant starts with the same question: “How can I help?”
Then comes the explaining. Your background. Your preferences. How you like things done. Again.
Rueyo is our answer: a small, personal AI model, fine-tuned on the experiences you choose to share. A model that learns your voice and carries your context into the next conversation.
One you. A world of possibilities.Your story is more than a chat history.
Rueyo is being designed to learn across it.
Messages, mail, photos, voice notes, and moments. Choose the sources that tell your story.
A small open-weight model, personally fine-tuned to your language, style, and patterns—with your corrections guiding it.
Rueyo is designed to collaborate with compatible agents, bringing the right understanding to the task at hand.
The product vision · Sources, model training, and agent integrations are in development.
See how a personal model could change
the everyday. Choose a moment.
“Really enjoyed that. Want to pick this up over coffee next week?”
An AI that knows you should answer to you.
These are the commitments shaping Rueyo.
We're designing for local inference on supported phones and laptops. Training requirements will be separate and clearly explained.
Choose what you connect, what Rueyo learns from, and what it shares with another agent.
People evolve. Your model and memory should be correctable, with a clear distinction between what you said and what AI inferred.
Understanding your preferences is different from permission to act. Rueyo should know the difference.
A FEW GOOD QUESTIONS
Rueyo is being built as your personal model and agent. You could interact with it directly, while compatible assistants could consult it for relevant context, preferences, or recommendations.
We start with a capable small open-weight model and adapt it using examples you choose to provide. The goal is to learn your expression and preferences. A separate memory system supports precise recall and changing facts.
That's the direction, but compatibility requires integrations. We're exploring standard interfaces such as MCP and agent-to-agent communication. No third-party integration is promised or available from this concept site.
Not yet. Rueyo is in development. This site introduces the vision and includes illustrative interactions; it does not train a model or connect to your personal accounts.
Local inference is a design goal. The exact storage, training, sync, and optional cloud processing choices are still being developed. We'll make those choices and their data flows clear before you connect anything.
We're building the personal thread that stays yours.
From one device, one assistant, one chapter to the next.