CIRIS - Your Privacy-First AI Assistant

CIRIS (Core Identity, Integrity, Resilience, Incompleteness, and Signalling Gratitude) is an ethical AI assistant that puts your privacy first. Unlike cloud-based AI apps, CIRIS runs its entire processing engine directly on your device.

🔒 Privacy by Design
Your conversations, memory, and data stay on YOUR device. The complete Python server runs locally - only LLM inference connects to the cloud. No data mining, no behavior tracking, no selling your information.

🤖 Ethical AI Framework
Built on CIRIS principles - an ethical AI architecture that prioritizes transparency, consent, and user autonomy. Every decision the AI makes follows a principled framework you can audit.

⚡ On-Device Processing
• Full FastAPI server runs on your phone
• SQLite database for secure local storage
• WebView UI for responsive interactions
• Works with any OpenAI-compatible LLM provider

🔐 Secure Authentication
• Google Sign-In for seamless account management
• JWT-based session security
• Role-based access control

💡 Key Features
• Natural conversation with AI assistant
• Memory system that remembers context
• Audit trail of all AI decisions
• Configurable LLM endpoints
• Consent management for data handling
• Dark/light theme support

📱 Technical Excellence
• Runs Python 3.10 via Chaquopy
• Supports ARM64, ARM32, and x86_64 devices
• Efficient memory usage (<500MB)
• Android 7.0+ compatible

💳 Credit System
Purchase credits through Google Play to power AI conversations. Your credits are tied to your Google account for easy management across devices. Credits are only required when using CIRIS proxied LLM services.

🌐 Bring Your Own LLM
Connect to any OpenAI-compatible endpoint - use OpenAI, Anthropic, local models, or self-hosted solutions. You control where your AI inference happens.

CIRIS represents a new approach to AI assistants: one that respects your privacy, operates transparently, and gives you control over your data and AI interactions.

https://github.com/cirisai/cirisagent