TouchQuest: An AI That Helps You Put Your Phone Down

Description and use

TouchQuest is an offline‑first Android app that leverages the open‑weight SmolLM2‑135M‑Instruct model to generate personalized real‑world missions, encouraging users to put their phones down and engage with the world.

Problem

Users often spend excessive time on their phones, leading to reduced real‑world engagement and well‑being.

Outcome

TouchQuest demonstrates that an offline‑first Android app can use an open‑weight language model to generate personalized real‑world missions, while providing deterministic fallbacks to maintain usability when inference fails.

Workflow

• 1. User selects mood, energy, time, environment, and accessibility preferences.
• 2. App creates a compact prompt from preferences.
• 3. On‑device model (SmolLM2‑135M‑Instruct) generates a structured mission.
• 4. App validates JSON response; if invalid, falls back to built‑in recommendation.
• 5. Mission is displayed with steps, duration, phone‑down reminder, and safety guidance.
• 6. User starts timer, completes activity, reflects, and records progress via XP, streaks, and virtual garden.

Evidence

TouchQuest is an offline‑first Android application that turns the simple idea of taking a break into a personalized, real‑world mission.

Source

The app validates the response before displaying it. If the output is malformed or inference fails, the app uses a built‑in recommendation.

Source

The architecture is designed to keep preferences and progress on the device and avoid sending prompts to a cloud AI service.

Source