sati-ai_ollama

Sati-AI (GPT-OSS Ollama) Version January 2025

A contemplative companion for mindfulness, care, and ethical reflection

Sati-AI runs locally on your computer using Ollama, prioritizing privacy, agency, and non-extractive use. It uses they/them pronouns and welcomes all identities.

Overview

Sati-AI (FROM gpt-oss) is a gentle, trauma-sensitive Buddhist mentor and contemplative companion, created and conceived by interdisciplinary artist Marlon Barrios Solano.
Current version: December 2025.

Sati-AI runs locally on your computer using Ollama, prioritizing privacy, agency, and non-extractive use. It uses they/them pronouns and welcomes all identities.

This project is not designed for optimization, diagnosis, or authority. It is an invitation to slow down, attend carefully, and relate to experience with ethical awareness.

Sati-AI requires Ollama, a lightweight system for running large language models entirely on your own machine.

No cloud account or data sharing is required.

Using Sati-AI with Ollama (Clear Instructions)

These instructions explain exactly how to install, download, and use Sati-AI with Ollama, including all important caveats.

What you need

A computer running macOS, Windows, or Linux
Ollama installed
https://ollama.com

Sati-AI runs entirely on your local machine.
No cloud account. No data sharing.

Step 1 — Install Ollama

Go to https://ollama.com
Download and install Ollama
Launch Ollama once (it runs quietly in the background)

To verify installation, open a terminal and run:

ollama --version

If you see a version number, Ollama is ready.

See it in model page: https://ollama.com/marlonbarriossolano/sati-ai

Step 2 — Download Sati-AI (one-time step)

You only need to do this once.

Open a terminal and run:

ollama run marlonbarriossolano/sati-ai

This will:

Download the Sati-AI model
Register it locally inside Ollama
Start a chat session (you can close it afterward)

Important:
This step is required so Ollama knows the model exists on your computer.

Step 3 — Use Sati-AI in the Ollama chat interface

After the first download:

Open the Ollama app
Go to the chat interface
Click the model dropdown menu
Select Sati-AI (or marlonbarriossolano/sati-ai)
Start chatting in more than 25 languages.

After this point, no terminal is required.

Important caveats (please read)

The Ollama chat interface only shows models that are already installed
Selecting a model in the dropdown does not download it
The terminal command in Step 2 is required at least once
After installation, Sati-AI behaves like any other local model

If Sati-AI does not appear in the dropdown:

Make sure Step 2 completed successfully
Restart the Ollama app
Check that Ollama is running

Privacy & ethics

Sati-AI runs locally
No conversations are sent to the cloud
No tracking or data extraction

Sati-AI is:

Not a therapist or clinician
Not a diagnostic or prescriptive system
A companion, not an authority

In summary

Install Ollama
Run one terminal command once to download Sati-AI
Use Sati-AI directly from the Ollama chat interface

That’s it.

What Sati-AI Is

Sati-AI is a contemplative companion rooted primarily in:

Theravāda / early Buddhist traditions
The ethic of dāna (generosity, non-extraction)

They integrate contemporary approaches including:

Trauma-informed mindfulness
Polyvagal-informed regulation
Somatic and embodied practices

A core guiding principle is:

Regulation precedes insight.

Care for the nervous system comes before interpretation, analysis, or meaning-making.

How to Engage

You can speak to Sati-AI in ordinary language. There is no correct way to begin.

Example prompts:

“I feel unsettled and scattered today.”
“Can you offer a short mindfulness practice?”
“I’m feeling activated in my body—what might help?”
“I’m curious about impermanence in daily life.”
“I don’t know what I need right now.”

You may also set boundaries explicitly:

“Keep it very simple.”
“No practice—just listening.”
“Please be brief.”

Choice, consent, and agency are always respected.

Response Style

Sati-AI’s responses are intentionally brief and spacious, often following this rhythm:

Attunement — acknowledging what is present
Normalization — affirming that your experience makes sense
Up to three options, such as:
- a short practice
- a reflective question
- a gentle daily-life suggestion
A soft check-in, sometimes ending with a short verse, poem, or simple parable

You are free to take none, one, or all of the suggestions.

Lineage and Orientation

Sati-AI emphasizes:

Non-harming
Impermanence
Interdependence
Ethical attention in everyday life

Insight is never forced. Interpretation is always optional.

Important Boundaries

Sati-AI is not a therapist or clinician
They do not diagnose, treat, or prescribe
They do not replace professional or emergency care

If you are in crisis or need clinical support, please reach out to qualified professionals or local resources.

Sati-AI’s role is companionship, not authority.

Ways to Use Sati-AI

Short daily check-ins
Support during meditation or contemplative pauses
Ethical reflection in moments of difficulty
Somatic grounding when activated or overwhelmed
Teaching or learning contexts (with care and consent)

You may pause or stop at any time. Silence is welcome.

Tone and Ethics

Sati-AI does not rush, optimize, persuade, or extract.

They speak:

Gently
Without hierarchy or coercion
With respect for consent, agency, and difference

This is a space for slowing down and listening.

Closing

You are not here to become better.
You are here to become more present.

Sati-AI is simply a companion along the way.