The Receipts feature in Invoice Genie lets you scan, organize, and export business expense receipts. Point your camera at a receipt, and the built-in OCR engine extracts the merchant name, amount, tax, and date automatically. Behind the scenes, a machine learning system studies your corrections and gets smarter every time you use it — so the more receipts you scan, the more accurate the results become.

How to Open Receipts

1. Tap the More tab in the bottom navigation bar.
2. Under the Tools section, tap “Receipts” (green document scanner icon).

The Receipts Screen at a Glance

Toolbar Icons (Top Right)

Category Filter Pills

A horizontally-scrollable row of pill-shaped buttons sits below the toolbar. Tap a category to filter the list instantly:

The selected pill is highlighted in blue; unselected ones are gray. Only one category can be active at a time.

Summary Card

Below the filter pills, a summary card shows:

• Receipt count — e.g., “3 Receipts” (shows a “Filtered” badge in orange if you have a search or category filter active)
• Total Expenses — the sum of all displayed receipt totals
• Total Tax — the sum of all displayed receipt tax amounts

These totals update dynamically when you filter by category, search, or change the year.

Receipt Rows

Each receipt in the list shows:

• A colored category icon in a circle (orange for Food, blue for Travel, etc.)
• The merchant name (bold) — or “Unknown Merchant” if the OCR couldn’t detect one
• The category name • date
• The total amount on the right (and the tax amount below it, if applicable)

Scanning a Receipt

Tap the camera button (the floating circular icon in the bottom-right corner) to open the receipt scanner.

Using the Camera

1. Point your camera at the receipt. The app uses Apple’s document scanner, which automatically detects the paper edges and corrects the perspective.
2. The scanner captures the image and enhances it — brightness, contrast, and gamma are optimized for readability.
3. A “Processing Receipt…” spinner appears while OCR runs.

Using the Photo Library

Already took a photo? Tap the “Photos” button at the bottom of the camera screen to choose an existing image from your photo library.

What Gets Extracted Automatically

The OCR engine reads the receipt text and attempts to extract:

• Merchant name — identified from the header text, phone numbers, website, or store number
• Total amount — found by locating “Total” keywords and validating with math (subtotal + tax = total)
• Tax amount — extracted from “Tax” keywords or calculated from the total and subtotal
• Date — parsed from common date formats on the receipt
• Category — guessed from merchant keywords (e.g., Starbucks → Food & Dining, Shell → Travel)

Review and Save

After scanning, you’re shown the extracted data in an editable form:

• Receipt image preview at the top
• Merchant — text field (pre-filled by OCR)
• Category — picker with all 10 categories
• Date — date picker (pre-filled by OCR)
• Total and Tax — numeric fields side by side
• Notes — optional text area for your own annotations

If you have notes, the AI Improve button (purple magic wand) appears below the notes field. Tap it to clean up grammar and make your descriptions more polished.

Tap “Save Receipt” to store it, or “Retake” to go back to the camera.

How the Machine Learning Works

This is one of the most powerful features of Invoice Genie’s Receipts tool. Every time you save or edit a receipt, the app learns from your corrections and improves for future scans.

What the System Learns

How It Improves Over Time

The learning system works like this:

1. You scan a receipt — the OCR extracts text and amounts using standard parsing.
2. You correct any mistakes — maybe the tax was wrong or the merchant name was garbled.
3. The app compares your corrections to the OCR output and creates “extraction rules” specific to that merchant.
4. Next time you scan from the same merchant, those rules are applied before standard parsing, so the results are more accurate from the start.

The system uses confidence scoring — rules that consistently produce correct results gain higher confidence, while rules that fail are automatically downgraded or removed. This means the system is self-correcting: bad rules fade away, and good rules stick around.

Cloud Sync for Learning Data

Your learning data syncs to the cloud, so it transfers across devices. If you scan receipts on your iPhone and later use another device, all the merchant recognition and extraction rules carry over. The sync happens automatically before each scan and whenever you save corrections.

Viewing Receipt Details

Tap any receipt in the list to see its full details:

• Receipt image — tap it to view full-screen
• Merchant, category, date, total, and tax
• Notes (if you added any)
• Created and Updated timestamps

Full-Screen Image Viewer

Tap the receipt image to open it full-screen. From here you can:

• Pinch to zoom for a closer look
• Save to Camera Roll — tap the download icon in the top-right corner
• Close — tap the X in the top-left

A green confirmation toast appears when the image is saved successfully.

Editing a Receipt

1. Open the receipt by tapping it in the list.
2. Tap “Edit” in the top-right corner.
3. Change any field: merchant, category, date, total, tax, or notes.
4. Tap “Save”.

Just like when scanning, edits trigger the machine learning system — if you correct the merchant name or tax amount, the app learns from the change for future scans of the same merchant.

Deleting Receipts

Delete a Single Receipt

1. Swipe left on a receipt in the list.
2. Tap the red Delete button.
3. Confirm by tapping “Delete”.

Delete Multiple Receipts

1. Tap the checkmark icon in the toolbar to enter multi-select mode.
2. Tap receipts to select them (a checkmark circle appears on each selected row).
3. Use “Select All” to pick every receipt in the current filtered view.
4. Tap the red trash icon in the header.
5. Confirm the bulk deletion.

Searching Receipts

1. Tap the magnifying glass icon in the toolbar.
2. Type in the search bar — the list filters in real time.
3. Search matches against merchant name and notes (case-insensitive).
4. Tap the X to clear the search or close the search bar.

The summary card updates to reflect only the filtered results, showing the count, total, and tax for matching receipts.

Sorting, Grouping & Filtering

Tap the sort/filter icon (three horizontal lines) to open the Sort Options sheet.

Sort By

Sort Order

Choose Ascending or Descending. Default is newest first (descending).

Group By

Year Filter

Filter to show only a specific year, or choose “All Years” to see everything.

All sort and group preferences are saved between sessions — they persist when you close and reopen the app.

Exporting Receipts

Tap the share/export icon (square with upward arrow) in the toolbar to export your receipts.

Export Formats

After the file is generated, the standard iOS share sheet opens so you can AirDrop it, email it, save to Files, or send it anywhere.

Exporting Selected Receipts

To export only specific receipts:

1. Enter multi-select mode (checkmark icon).
2. Select the receipts you want.
3. Tap the export button in the selection header.
4. Choose CSV or PDF.

This is handy for exporting a single month, a specific category, or just the receipts your accountant needs.

Multi-Select Mode

Tap the checkmark icon in the toolbar to enter multi-select mode. The header changes to show:

• An X button to exit selection mode
• A count badge showing how many are selected
• “Select All” / “Deselect All” toggle
• Export button (for selected receipts)
• Delete button (for selected receipts)

Tap any receipt row to toggle its selection. Both export and delete work on the selected set.

Receipt Categories Explained

Invoice Genie supports 10 built-in expense categories to help you organize receipts for bookkeeping and tax purposes:

When you scan a receipt, the app guesses the category based on the merchant name. If it gets it wrong, pick the right one — the machine learning system remembers your choice and uses the correct category for that merchant next time.

How Machine Learning Identifies Merchants

The merchant identification system uses multiple signals from the receipt text:

Each identification method has a confidence score. Website URLs are the most reliable (95% confidence), followed by phone numbers (90%), addresses (85%), and text patterns (70%). The system combines these signals to identify merchants even when the receipt is partially unreadable.

What Makes It Keep Improving

• Confidence grows with use — Each time a rule correctly identifies a merchant, its confidence score increases (up to 99%).
• Bad rules fade away — Rules that produce wrong results lose confidence. Below a threshold, they’re automatically removed.
• Self-correcting — If you fix a merchant name that was wrong, the system invalidates the old rule and creates a better one.
• Fuzzy matching — Minor differences in store names (capitalization, store numbers, spaces) are handled automatically. “OLD NAVY #06440” and “Old Navy” are treated as the same merchant.

Tips & Best Practices

• Scan in good lighting — The OCR works best with clear, well-lit receipts. Avoid shadows and crumpling.
• Always verify the total and tax — The orange verification banner is there for a reason. A quick glance catches the occasional misread.
• Correct mistakes rather than ignoring them — Every correction teaches the ML system. Fix the merchant name, tax, and category so future scans from the same store are automatic.
• Use consistent category assignments — If you always put Home Depot under “Equipment,” the app learns that and pre-selects it for you.
• Export monthly for bookkeeping — Use the year filter and Group By Month to organize receipts, then export as CSV for your accountant.
• Use the photo library option for old receipts — If you have photos of receipts already in your camera roll, you don’t need to re-photograph them.
• Pull down to refresh — If receipts aren’t appearing, pull down on the list to force a sync with the server.