Google Sheets Add-On

Parsie Advanced OCR is a powerful Google Sheets add-on that transforms PDF documents, images, spreadsheets, and Word documents into structured data in Google Sheets. Using Azure's industry-leading OCR algorithms, Parsie can extract text from both handwritten and computer-generated documents with exceptional accuracy.

Unlike traditional OCR tools that simply convert documents to text and preserve their original messy layout, Parsie uses a template-first approach:

1. You define the data structure you need - Create a schema that specifies exactly what information matters to your business
2. Parsie extracts only what you need - Our AI identifies and pulls just the relevant data points
3. Get consistent, structured output every time - No matter what format your input documents use

Example: Insurance professionals processing loss run reports from multiple carriers, each with wildly different layouts and formats. With Parsie, you define the key data points once (claim numbers, loss dates, reserves, etc.), and get uniform, structured data ready for analysis - regardless of which carrier format you're processing.

Parsie works with virtually any document type you need to process:

• PDF Documents

  • Digital native PDFs
  • Scanned documents
  • Forms and structured documents

• Images

  • JPG/JPEG
  • PNG
  • Screenshots

• Microsoft Office Documents

  • Word (.doc, .docx)
  • Excel (.xls, .xlsx)

Parsie has these Text Recognition Capabilities built-in:

• Machine-printed text
• Handwritten text
• Tables and structured data
• Multi-page documents

Follow these steps to get started with Parsie Google Sheets Add-on:

1. Install directly from Google Workspace Marketplace
2. Test basic functionalities anonymously with 3 free pages
3. Sign up and create an account with Parsie directly from Google Sheets to receive 20 pages free trial with advanced functionalities

When you open a new Sheet, follow these steps:

1. Notice the yellow notification saying "table not configured" in the side panel
2. Rename your sheet to be descriptive (this will become the name of the table)
3. Click the gear icon "Configure" to set up your table

Table Configuration Options

You can either:

• Import structure/schema from an existing table in Parsie (you have to sign in first)
• Create a new table structure/schema from scratch

Creating a New Table in Google Sheets

1. Click "Add column" and define each column data type (text, number, yes/no)
2. Use descriptive column names
3. Optionally add data descriptions and example values to increase extraction accuracy
4. Click "Save and sync with Parsie" when finished

Extracting Multiple Rows from a Single Document

By default, Parsie Google Sheets add-on extracts one row of data per document. For more complex documents like:

• Bank statements with multiple transactions
• Insurance loss run reports with multiple claims

You'll need to configure a template from the Parsie webapp then connect your Sheet with that template.

Advanced Configuration via Parsie Dashboard

1. Navigate to your Parsie dashboard
2. Go to the Table tab
3. Click "New Table"

Choose from these options:

• Use a ready-made template (resume, loss run, etc.)
• Use AI to analyze a sample document and generate a template
• Create a template from scratch

Creating a Custom Template

When creating columns, configure:

• Column Name (Required)

  • Descriptive name (e.g., "Product Name")

• Column Type (Required)

  • String → Text (e.g., "Product Name" → "iPhone 15")
  • Number → Numeric values (e.g., "Price" → 999.99)
  • Boolean → True/False (e.g., "In Stock" → true)
  • Object → Structured values with multiple fields

• Additional Configuration

  • Optional fields: Add descriptions or examples to improve AI extraction
  • Lists: Enable for columns containing multiple values

Example: Receipt with Multiple Line Items

To parse receipts with multiple items per document:

1. Create a "Line item" column with:
2. Data type: Object
3. Enable "List" checkbox

Line Item Configuration

Add these properties to the Line item object:

• Required Fields

  • Name: text (e.g., "Apple Pie")
  • Unit: number (e.g., piece/kg)
  • Unit Price: number (e.g., 10)
  • Currency: text (e.g., $)
  • Total: number

When exported, the data will be flattened with one row per line item, with shared values (like store name) repeated across related rows.

Once you are done with table configuration, Parsie offers two methods for uploading documents:

• Local files

  • Drag/drop or select files from your computer

• Google Drive

  • Get the sharing URL to your files or folders
  • Set permissions to "anyone with the link can view"
  • Provide multiple links separated by commas
  • You can change permissions back to private after processing

After uploading, click "Extract data" and wait for your sheet to populate with the extracted information.

All uploaded files are deleted from our servers once processing is complete.

For questions or improvement suggestions, contact us or email help@parsie.pro.