Capital Shack — Document OCR Processing
Shipped Jun – Jul 2025
Capital Shack's client documents — application forms, bank statements, and other supporting paperwork — arrived by email in a continuous stream, and each one had to be opened, categorized, filed, and cross-referenced into a spreadsheet by hand before it could be used downstream. That manual triage held up at low volume but became a slow, repetitive bottleneck — and a source of transcription errors — as volume grew.
Client confidentiality prevents sharing screenshots or product media for this project.
Built With
Gmail 
Google Drive 
Google Sheets 
Google Apps Script 
Google Form Parser (Document AI) Project Overview
Capital Shack's client documents — application forms, bank statements, and other supporting paperwork — arrived by email in a continuous stream, and each one had to be opened, categorized, filed, and cross-referenced into a spreadsheet by hand before it could be used downstream. That manual triage held up at low volume but became a slow, repetitive bottleneck — and a source of transcription errors — as volume grew.
Two linked Google Apps Script automations now handle the pipeline end to end. The first reviews each email as it arrives, categorizes it, uploads its attachments into a specifically labeled Google Drive folder, and logs an unprocessed entry in Google Sheets; the second picks up those entries and runs the attachments through Google Form Parser (Document AI), identifying whether each is an application form, bank statement, or other document type and writing the extracted data into the correct sheet fields automatically — removing manual triage and data entry from the workflow entirely.
- Challenge
- Client documents arrived by email in a continuous, unpredictable stream — application forms, bank statements, and other supporting paperwork — and every one had to be opened, read, categorized, filed into the right place, and cross-referenced into a spreadsheet by hand before anyone downstream could act on it. That process held up fine at low volume, but became a genuine bottleneck as volume grew: slow, repetitive, and prone to the transcription and filing mistakes that come from doing the same manual triage over and over.
- Solution
- Built two linked automations in Google Apps Script that hand off to each other through a shared Google Sheet. The first watches the inbox as documents arrive: it reviews each email's attachments, categorizes the message, uploads the attachments into a specifically labeled Google Drive folder, and logs a corresponding entry in Google Sheets marked unprocessed. The second automation picks up those newly logged, unprocessed entries on its own schedule and runs each attachment through Google Form Parser (Document AI), which identifies whether it's an application form, a bank statement, or another document type and extracts the relevant structured data — writing the results into the correct fields in the appropriate sheet automatically.
- Impact
- Manual document triage and data entry were eliminated entirely: every incoming document is now consistently categorized, filed into Drive, and logged and populated in Sheets without anyone needing to open or process it by hand — removing both the bottleneck and the transcription errors that came with doing it manually.
Frequently Asked Questions
Manually reviewing, categorizing, filing, and cross-referencing every incoming client document into spreadsheets was slow, repetitive, and error-prone at any real volume.
Capital Shack — Document OCR Processing was built with Gmail, Google Drive, Google Sheets, Google Apps Script, Google Form Parser (Document AI).
No — this was built under client confidentiality, so neither the source code nor product details beyond this summary are publicly shareable.
An Apps Script trigger runs on a schedule, scanning the inbox via Gmail's Apps Script service for new messages, reading each one's attachments, categorizing the message, and taking an action — like saving the attachment to a specific Drive folder and logging an entry in Sheets — without a human opening a single email.
Google Form Parser (Document AI) is Google's OCR/document-understanding service — pointed at a scanned or PDF document, it reads and structures the text into labeled fields, which is what lets the second automation tell an application form apart from a bank statement and extract the right data from each without a hardcoded template per document type.
Splitting the pipeline lets each automation run on its own trigger: intake reacts to new email as it arrives, while OCR processing works through everything currently marked unprocessed in the shared sheet on its own schedule — so a slower Document AI call never blocks or delays new documents from being filed and logged.
Available for work
Got a project like this
in mind?
Let's talk about what you're building and how I can help ship it.