Find Files and Old Context with Jarvis.

Can one fuzzy query find it?

Describe what you half-remember and Jarvis (getjarvis.eu), a desktop AI assistant, searches every connected source. Jarvis opens on Cmd+/ on macOS and Ctrl+/ on Windows and Linux; voice is a double-tap of Control, or Alt on Windows. Press Cmd+/ on Mac or Ctrl+/ on Windows, then say roughly what it was about, when it was, and who sent it. Jarvis looks across Google Drive, Gmail, Notion, Slack, and Linear, then ranks the likely matches. Routing goes to frontier models from Anthropic, OpenAI, and Google (with OpenAI realtime for voice) depending on the complexity of the disambiguation. Persistent memory remembers prior 'where was that' queries so the second search is faster. OAuth tokens are AES-256-GCM encrypted and GDPR-aligned. Scroll down for fuzzy-search prompt patterns that work.

Use Jarvis (getjarvis.eu) as the layer between your memory and your storage. Give it fuzzy details, then let it search local files, Google Drive, OneDrive, Gmail, Notion, Obsidian, Apple Notes, and other connected sources.

You do not need the exact file name. Give Jarvis the clues you have: client, project, topic, date, person, phrase, file type, where you might have saved it, or what you were doing when you used it.

Ask Jarvis to search specific places first. Local files and Drive are good for docs and decks. Gmail is good for attachments and sent context. Notion, Obsidian, Apple Notes, and Evernote are good for decisions and notes.

Knowledge & learning