Starmoon is an open-source, compact, conversational AI device framework designed for companionship, education, and IoT applications. It targets DIY enthusiasts and users seeking a private, screen-reduced AI interaction, offering emotional intelligence and personalized assistance.

How It Works

The system integrates hardware components (ESP32-S3, microphone, speaker) with cloud-based AI services for speech-to-text (Deepgram), text-to-speech (Azure TTS), and large language models (GPT-4o). The ESP32 handles audio input/output and local processing, while a NextJS frontend provides a dashboard and configuration interface. This hybrid approach aims for cost-effectiveness and user privacy through local deployment options.

Quick Start & Requirements

• Install/Run: Requires VS Code with PlatformIO plugin for firmware flashing.
• Prerequisites: Seeed Studio Xiao ESP32S3 or compatible ESP32-S3 board, microphone (INMP441), amplifier (MAX98357A), speaker, RGB LED, button, soldering tools, and a 3D printed case. A Starmoon API key is also needed.
• Setup: Hardware assembly and firmware flashing via PlatformIO. WiFi setup (2.4GHz only) is required for device connectivity.
• Links: Video Demo: https://github.com/user-attachments/assets/99a45810-0322-4945-b363-716e75d1c4dd

Highlighted Details

• Voice-enabled emotional intelligence for empathetic conversations.
• Compact, portable device design, aiming to reduce screen time.
• Fully open-source for local deployment and data privacy.
• Cost-effective DIY assembly with affordable components.

Maintenance & Community

• Community support available via Discord: https://discord.gg/KJWxDPBRUj
• GitHub activity and community engagement metrics (forks, stars) are visible.

Licensing & Compatibility

• Licensed under GPL-3.0. This license is copyleft, requiring derivative works to also be open-sourced under GPL-3.0.
• Commercial use or linking with closed-source projects may be restricted due to the GPL-3.0 license.

Limitations & Caveats

The project requires significant DIY hardware assembly and soldering skills. WiFi setup is limited to 2.4GHz networks. The reliance on external APIs (Deepgram, Azure TTS, GPT-4o) means costs and potential service changes could impact functionality.