I just launched a Google Sheets Add-on called Parsie, which helps you extract structured data and tables (like names, emails, invoice totals, etc.) from unstructured documents directly in your Google Sheets — including PDFs, screenshots, and more.
Unlike basic OCR tools that just dump messy text, Parsie understands documents like a human would. It uses a template-first approach:
1) You define what data you need
2) Parsie extracts only that
3) You get clean, consistent output.
Under the hood:
– Powered by GPT models + Microsoft Azure OCR (top-ranked since 2018)
– Understands context and relationships between data points
– Works in 100+ languages
– Handles scanned PDFs, images, DOCX, handwriting, and complex layouts
Use cases:
– Invoices, receipts, and bank statements
– Insurance and legal docs
– Form submissions
– Any workflow that turns messy documents into structured data
Advanced features:
– AI-assisted custom schema
– Multi-row extraction
– Batch document processing
– Metadata (file name, Drive URL, etc.)
Try it here: https://workspace.google.com/marketplace/app/advanced_ocr_ex...
Website: https://parsie.pro/
Would love your feedback or ideas for improvement. AMA!
