DocuVo
DocuVo reads invoices and delivery notes, extracts number and date, renames the files correctly and delivers them import-ready as PDF, JPG or TIFF – as a multi-tenant SaaS for teams in Austria and Germany.
- Role
- Product, backend, design
- Status
- Live
- Stack
- Python · Flask · Tesseract · OpenCV · spaCy · Celery
- Website
- View live · docuvuu.com
- Automated tests
- 205
- Export
- PDF · JPG · TIFF
- Multi-tenant
- Yes
01Challenge
Documents arrive as PDF, scan or photo – in every layout. Typing and renaming them costs time and causes errors before accounting even starts.
02Solution
A recognition pipeline of OCR, image processing and a language-model fallback (scans included), with templates per document type, confidence scoring and batch review for uncertain cases. Teams work in their own organisations with roles and branding; processing runs asynchronously in the background.
03Outcome
Live at docuvuu.com – backed by 205 automated tests, CI and GDPR-compliant auto-deletion.
Glimpses
Next project
Memecoin Intelligence
Research and risk radar for Solana memecoins.