Spaces:
Running
Running
File size: 4,841 Bytes
9d8ea9c c5fee9e 9d8ea9c c5fee9e 9d8ea9c c5fee9e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 |
---
title: Echo
emoji: "π€"
colorFrom: indigo
colorTo: purple
sdk: static
pinned: false
license: mit
short_description: "A companion robot that remembers you and grows with you"
tags:
- reachy_mini
- reachy_mini_python_app
- companion
- memory
- ai
---
# Reachy Echo
A companion robot that remembers you and grows with you.
## What Makes Echo Different
Most robot apps treat the robot as a voice interface with decorative movement. Echo is different:
| Feature | Traditional | Echo |
|---------|-------------|------|
| **Memory** | Forgets everything | Remembers your name, preferences, conversations |
| **Initiative** | Waits for commands | Greets you, suggests breaks, celebrates wins |
| **Models** | Single provider | 18+ models via LiteLLM (swap anytime) |
| **Movement** | Decorative | Communicates emotion and state |
## Quick Start
### Install
```bash
cd ~/apps/reachy/apps/echo
pip install -e .
```
### Run (Simulation Mode)
```bash
python -m reachy_mini_echo --sim
```
Open http://localhost:7861
### Run (Real Robot)
With the Reachy daemon running:
- Echo appears in the Reachy Mini dashboard
- Or access directly at http://localhost:7861
## Features
### Memory System
Echo remembers across sessions:
- Your name and preferences
- Past conversations and topics
- Work patterns and habits
Try saying:
- "My name is Alex"
- "I'm a software engineer"
- "What do you know about me?"
### Proactive Behaviors
| Behavior | What it does | When |
|----------|--------------|------|
| **Morning Greeting** | Personalized hello | First appearance, 6-11am |
| **Work Break Reminder** | Suggests a stretch | After 2 hours of work |
| **Build Celebration** | Excited dance | When your code builds |
| **Build Support** | Sympathetic response | When builds fail |
| **Return Greeting** | Welcome back | After 30+ min absence |
### Model Selection
Switch between 18+ models instantly:
| Model | Best for |
|-------|----------|
| `llama-3.3-70b-cerebras` | Fast local inference (default) |
| `claude-opus-4.5-openrouter` | Most capable |
| `gpt-5.2-openrouter` | Latest GPT |
| `gemini-3-pro-openrouter` | Multimodal |
| `qwen-3-235b-cerebras` | Large context |
## Configuration
Environment variables (optional):
```bash
# LiteLLM server (defaults to DGX)
export LITELLM_URL=http://your-litellm-server:4000
export LITELLM_API_KEY=your-key
# Default model
export LITELLM_MODEL=llama-3.3-70b-cerebras
```
## UI Overview
```
βββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β π€ Reachy Echo β
β A companion that remembers you and grows with you β
βββββββββββββββββββββββββββββββ¬ββββββββββββββββββββββββ€
β β Status: π’ Connected β
β Conversation β β
β βββββββββββββ β Model: [dropdown] β
β User: Hi there! β β
β Echo: Hello! How are you? β Memory: β
β β Facts: 3 | Sessions: 5β
β β β
β [Type a message...] [Send]β Proactive Behaviors: β
β β β Morning Greeting β
β β β Work Break Reminder β
β β β Build Celebration β
β β β
β β [Clear] [Forget Me] β
βββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββ
```
## Architecture
```
reachy_mini_echo/
βββ main.py # ReachyMiniEcho app class
βββ providers/ # LLM backends (LiteLLM)
βββ memory/ # SQLite + fact extraction
βββ proactive/ # Trigger/behavior engine
```
## Development
See `CLAUDE.md` for detailed architecture and development guide.
### Adding Behaviors
1. Create behavior class in `proactive/behaviors.py`
2. Implement `execute(echo)` method
3. Create trigger in `proactive/triggers.py`
4. Register in `proactive/engine.py`
### Memory Fact Extraction
Echo automatically extracts facts from conversation:
- "My name is X" β stores name
- "I prefer X" β stores preference
- "I work at X" β stores employer
## Privacy
- All data stored locally in `data/memory.db`
- "Forget Me" button clears personal data
- No cloud storage of conversations
- Local models available for full privacy
## License
MIT
|