Loading...
Not Loading? Load a simpler page!
← Back to Projects
Safety & AI Utility
AerisIQ
Privacy-first disaster risk intelligence and local safety utility for Android, featuring an offline LLM.
Overview
AerisIQ is a Free and Open Source Software (FOSS), privacy-first disaster risk intelligence and local safety utility application for Android. The app queries public warning feeds, parses raw disaster bulletins, combines them with live local weather telemetry, and processes the combined datasets using an offline Large Language Model (LLM) running securely in the device's sandbox.
Why It Exists
Built to operate under strict user-sovereignty principles. 100% offline AI ensures safety summaries are executed client-side with no chat logs transmitted. It uses a zero-telemetry approach, meaning no analytical trackers or advertisement loggers.
Key Features
100% Offline AI
Safety summaries are executed client-side using Qwen 2.5 0.5B Instruct model via MediaPipe LLM Inference APIs.
Geocoding API
Coordinates are resolved via client-side API requests. If location permissions are withheld, users can select regions manually.
Safety Lock
For extreme warnings, a 6-step checklist card is displayed. Users must check off all procedures before AI analysis is unlocked.
Premium UI
Features a premium dark glassmorphic design system using Google Sans Flex variable-axis typography and directional page-slide transitions.
Zero Telemetry
No trackers are present. Clearing app storage permanently deletes all local files and AI weights.
