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
Chain-of-Thought Reasoning
Step-by-step reasoning prompt for complex problem solving
Few-Shot Learning Prompt Builder
Create effective few-shot learning prompts with examples that teach AI models new tasks quickly.
System Design Evaluator
Evaluate and improve system architecture designs
Data Pipeline Architecture Review
Review and optimize data pipeline architectures for performance and reliability