Invoice numbers need keyword boost; clause meaning needs embeddings. DigUp combines both with on-device EmbeddingGemma 2 (~865 MB once). Not a cloud OCR SaaS — community docs only.
Index finance PDF folders; natural query + exact ID
Exact-word boost for invoice numbers (README)
Pet clause in a lease PDF
Page-level PDF embeddings
Open at the page — classic DigUp demo
Bank statement screenshots
Image embeddings (+ Mac OCR)
Search “March payroll remittance” by look/text
Audio from a tax call
30s audio windows
Jump to the discussion window
Numbered checklist
Install DigUp; verify version (Mac 0.6.1 / Win·Linux beta).
Index encrypted or cloud-only vaults only after files are local (DigUp skips online-only).
Prefer PDF/docx folders DigUp reads; xlsx/pptx are limited upstream (#75) — export PDF when needed.
Search meaning + paste exact invoice IDs.
Read privacy notes: no DigUp account; optional update check.
Why EmbeddingGemma 2 matters here
Google’s EmbeddingGemma 2 (~740M, unified 768-d multimodal space) is what DigUp runs via llama.cpp (~865 MB Q8 package). One query can rank images, PDF pages, audio windows, video frames, and optional code — see the local multimodal guide and model page. digup.run is not Google-official.
DigUp