Summary

LifeTrace is an AI-powered system designed to automatically manage personal task context by recording, organizing, and retrieving daily activity traces. It targets users seeking enhanced productivity and information recall through intelligent life logging, offering a novel approach to personal data management.

How It Works

The system employs timed automatic screenshots and OCR (RapidOCR) to capture and digitize user activities. These are aggregated into "intelligent events" based on context, with vector retrieval and multimodal search enabling efficient past information retrieval. This approach automates the organization of digital life traces, transforming raw activity data into actionable insights.

Quick Start & Requirements

• Requires Python 3.13+ and is supported on Windows and macOS. Optional CUDA support is available for GPU acceleration.
• Dependency management uses uv. Installation involves installing uv, syncing dependencies (uv sync), activating a virtual environment, and configuring settings by copying default_config.yaml to config.yaml.
• The backend service is started via python -m lifetrace.server (at http://localhost:8000), and the frontend via cd frontend && pnpm install && pnpm dev (at http://localhost:3000).
• Links: Documentation, Quick Start, Features, Development, Contributing (details within README).

Highlighted Details

• Automatic screenshot recording and OCR for activity capture.
• AI-driven aggregation of screenshots into contextual "intelligent events."
• Time allocation analysis with app categorization and 24-hour visualizations.
• Comprehensive RESTful Web API service for integration.
• Multimodal search capabilities for retrieving past information fragments.

Maintenance & Community

• Community engagement is encouraged via WeChat, Feishu, and Xiaohongshu groups.
• Contributions are welcomed through pull requests, with specific areas like bug reports, feature requests, documentation, and code implementation.
• Documentation is managed via deepwiki.

Licensing & Compatibility

• The software is licensed under the Apache License 2.0, which permits commercial use and integration into closed-source projects. Learning resources are copyright © 2025 LifeTrace.org.

Limitations & Caveats

The README does not specify alpha/beta status or known limitations. CUDA is optional, implying potential performance differences between GPU and CPU execution for AI tasks. Both backend and frontend services must be actively running for full functionality.