How It Works
V1 Pipeline
Upload files
↓ parse + extract text
↓ chunk + tag metadata
↓ index (persistent, this browser)
Ask question
↓ retrieve relevant excerpts
↓ AI analysis with citations
Generate report
↓ markdown / copy / download
Documents are processed and stored locally in this browser (IndexedDB).
Nothing is uploaded to any server. With an API key configured, document
excerpts are sent directly to the Anthropic API for analysis.