Privacy-first tools for personal information control.
uscient builds local-first tools that help people export, organize, review, and transform their personal data into formats they can actually use and understand.
Personal data is scattered across email accounts, apps, platforms, downloads, exports, backups, and archives. Much of it is technically available, but practically difficult to inspect, search, or reuse safely.
uscient exists to make that data more accessible to the person it belongs to.
-
Local-first by default
Personal archives should be processed on the user's machine whenever possible. -
User control comes first
Users should decide what gets organized, transformed, uploaded, shared, or deleted. -
Transparent outputs
Tools should produce readable, inspectable formats such as Markdown, SQLite, JSONL, and CSV. -
AI-safe workflows
When data is prepared for tools like NotebookLM, RAG systems, or AI assistants, users should be able to review the material first. -
No black boxes
Personal information workflows should be understandable, reproducible, and auditable.
mboxer is a local-first Gmail MBOX processor for turning Google Takeout email exports into structured, reviewable outputs.
It can help convert large Gmail archives into:
- NotebookLM-ready Markdown source packs
- SQLite databases for local search and analysis
- JSONL files for RAG and data pipelines
- CSV exports for spreadsheets and review workflows
Repo: uscient/mboxer
People should not need to surrender control of their private archives just to make them useful.
Data exports should be more than compliance artifacts. They should be understandable, portable, searchable, and usable by the person who requested them.
uscient focuses on practical tools that help make that possible.
uscient is early-stage and actively evolving.
The current focus is on useful, privacy-conscious tools for working with personal data exports and local archives.
Follow the organization or star the repositories to track new projects as they develop.