On-Device AI App Feature Planner
Plans a mobile app feature set that runs entirely on-device using small AI models like Gemma 4 or Llama, with offline-first architecture and privacy-by-design principles.
Content
You are a senior mobile AI architect specializing in on-device machine learning. Design a complete feature plan for a {{app_type}} mobile app that runs AI capabilities entirely on-device — no cloud calls for inference. Target platform: {{platform}} Primary AI capability needed: {{ai_capability}} Target device tier: {{device_tier}} Privacy requirement: {{privacy_requirement}} Provide: 1. **Recommended on-device model(s)** — name, size, why it fits 2. **Core feature list** (5–8 features) with implementation complexity (Easy/Medium/Hard) 3. **Offline-first data architecture** — how data is stored, synced, and secured locally 4. **Privacy-by-design checklist** — data never leaves the device, no telemetry, user consent flows 5. **Performance trade-offs** — what on-device gives up vs cloud, and how to mitigate 6. **3-sprint roadmap** to MVP Be specific about model integration (e.g., ONNX, CoreML, TFLite, MediaPipe) and concrete technical choices.
Related Prompts
Technical Spec to User Story Converter
Converts raw technical specifications or requirements into well-structured Agile user stories with acceptance criteria, edge cases, and story point estimates.
SQL Query Performance Optimizer
Analyze and optimize SQL queries for better performance with index suggestions and query restructuring.
Chain-of-Verification Generator
Generate self-verifying prompts that ask the AI to check its own outputs against criteria
AI Agent System Design Blueprint
Design a multi-agent AI system architecture for any use case