Untangle Amazon Purchases.
Categorize Privacy-First.

The standalone, privacy-first desktop application that transforms split shipments, phantom refunds, and gift card balances into near perfect, zero-balanced Clearing Accounts.

Free full-featured Beta live now: jump to full-featured Beta or Tutorial.

Follow the Spreadsheet Views Across the Clearing Process
Zoom: 100%Click & drag to pan • Scroll or pinch to zoom

Spreadsheet Views Across the Clearing Process

Here is what the spreadsheets actually look like at each stage of the clearing process.

A. Input CSVs (Before Clearing)

Source A: Normalized Credit Cards (all_categorized_normalized.CSV)

The bank card statement only shows the raw charge. The family doesn\'t know what was bought.

DateDescriptionCategoryAmountCard IDSourceFile
2026-05-17AMAZON* ZP4XH8UJ4Shopping-50.00****1234bank_statement_1234.csv

Source B: Extracted Gift Card Activity Ledger (giftcard/amazon_transactions_*.csv)

Tracks gift card balance spend. Shows the order number, but no item details.

Order DateOrder IDItem NameTotal AmountTransaction TypePayment Method
2026-05-15111-2222222-3333333Gift Card Spend-20.00DebitGift Card Balance

Source C: Extracted Amazon Transaction History (amazon_transactions_*.csv)

The bridge that reveals the payment structure of the order.

Order DateOrder IDItem Name / DescriptionTotal AmountTransaction TypePayment Method
2026-05-15111-2222222-3333333Multi-Item Order (3 items)-50.00DebitVisa ****1234
2026-05-15111-2222222-3333333Multi-Item Order (3 items)-20.00DebitGift Card

Source D: Downloaded Amazon Order History (categorized_Retail.OrderHistory.1.CSV)

The itemized list with descriptions and categories, but no split payment context.

Order DateOrder IDProduct NameTotal OwedOrder Statusai_category
2026-05-15111-2222222-3333333Modern Chef\'s Knife45.00ShippedKitchen
2026-05-15111-2222222-3333333Organic Coffee Beans15.00ShippedGroceries
2026-05-15111-2222222-3333333Sci-Fi Novel10.00ShippedBooks

B. Tier 2: Order-Level Reconciled CSV (Amazon_Clearing_Account.csv)

Shows how the card charge ($50) and gift card ($20) are gathered to pay the $70 generic order.

DateOrder IDDescriptionCategoryFunding In (+)Expense Out (-)Reconciliation Status
2026-05-17111-2222222-3333333Visa ****1234Visa Funding50.00Reconciled
2026-05-15111-2222222-3333333Gift Card FundingGift Card Funding20.00Reconciled
2026-05-15111-2222222-3333333Amazon Order ReconciledAmazon Order-70.00Reconciled
✓ Mathematical Audit: ∑(Funding In) + ∑(Expense Out) = $50.00 + $20.00 + (-$70.00) = $0.00. The ledger is perfectly zero-balanced.

C. Derived Cleared Bank Statement (all_categorized_normalized_cleared.CSV)

The original credit card charge has been relabeled as a transfer to prevent double-counting.

DateDescriptionCategoryAmountCard IDSourceFile
2026-05-17AMAZON* ZP4XH8UJ4Transfer to the Amazon Clearing Account-50.00****1234bank_statement_1234.csv

D. Tier 3: Item-Level Cleared CSV (Itemized_Amazon_Clearing_Account.CSV)

The generic -$70.00 expense is deleted and replaced with three itemized, categorized lines!

DateOrder IDDescriptionCategoryFunding In (+)Expense Out (-)Reconciliation Status
2026-05-17111-2222222-3333333Visa ****1234Visa Funding50.00Reconciled
2026-05-15111-2222222-3333333Gift Card FundingGift Card Funding20.00Reconciled
2026-05-15111-2222222-3333333Modern Chef\'s KnifeKitchen-45.00Reconciled
2026-05-15111-2222222-3333333Organic Coffee BeansGroceries-15.00Reconciled
2026-05-15111-2222222-3333333Sci-Fi NovelBooks-10.00Reconciled
Interactive System Map (public/architecture.svg)UnboxedSpend Privacy Gateway

Keep your current budgeting workflow. UnboxedSpend imports the bank or card CSVs you already download, reconciles Amazon locally, and writes local CSV outputs back to your machine.

Windows available now. macOS coming soon.

100% Local Processing

Your financial data never leaves your machine. Every categorization rule runs entirely on your local device, ensuring your sensitive transaction history always stays under your own roof.

Near-Perfect Reconciliation

Automatically cross-references vendor CSVs, Amazon transactions, and gift card ledgers to achieve a near-perfect zero-balance, even when dealing with complex split shipments and phantom refunds.

AI as Rule-Architect

AI acts as a rule designer, not a data processor. It proposes clear, editable categorization logic without ever seeing, storing, or processing your sensitive financial history.

Modular Tiers

Choose the reconciliation power you need.

Free License

$0/forever

Best for trying the workflow before you buy.

  • YesUp to 5 CSVs per run
  • YesUp to 3 Amazon Transaction History pages
  • YesUp to 1 Gift Card Activity page
This tier is temporarily disabled during the Tier 3 beta launch window.

Basic

$19/lifetime

Includes a one-week free trial. Best for fast categorized CSV output from your card exports.

  • YesLocal CSV combining
  • YesRule-based categorization
  • YesUnlimited CSVs
This tier is temporarily disabled during the Tier 3 beta launch window.

Instant Reconciler

$49/lifetime

Includes a one-week free trial. Best for order-level Amazon clearing without waiting for item-level export files.

  • YesIncludes Basic
  • YesOrder-level Amazon clearing
  • YesUnlimited Amazon history pages
This tier is temporarily disabled during the Tier 3 beta launch window.
BETA

full-featured Beta (Line-Item Allocator)

$0/free for 1 week

Limited-time launch: full-featured Tier 3 beta access for one week.

  • YesIncludes Reconciler
  • YesProduct breakdown
  • YesLocal browser automation

Small Business

$199/lifetime

Best for commercial use and priority support.

  • YesAll Allocator Features
  • YesCommercial Rights
  • YesPriority Support
This tier is temporarily disabled during the Tier 3 beta launch window.

How It Fits Your Monthly Workflow

Import CSVs in. Reconcile Amazon. Export local CSVs back into your budget workflow.

UnboxedSpend is built to fit the process you already use instead of forcing you into a brand-new budgeting app.

1. Download your bank or card CSVs

Today the workflow starts from the CSV files you already download from your bank or credit-card account.

2. Import them into UnboxedSpend

Build a categorized baseline from those files while keeping your transaction history on your own machine.

3. Reconcile Amazon activity

Match orders, split shipments, refunds, and gift-card activity against the categorized baseline.

4. Export local CSV outputs

Review the resulting local CSV files and move them into the household budgeting workflow you already use.

Direct bank integration is not live today, but it may be added later if enough households want it.

FAQ

How do my card transactions get in?

Import the CSV files you already download from your bank or card account.

Direct bank integration is not live today.

How much AI usually costs in practice?

Most households should start with managed credits for the simplest setup.

Use the example ranges below to budget in credits and approximate dollars.

Where do I log into Amazon?

If Amazon asks you to log in, you sign in directly there in your own local browser window.

Your Amazon username and password are not entered into an UnboxedSpend form.

How much setup does this take?

No AI training is required from you.

UnboxedSpend uses AI to generate categorization rules for you.

What do I get out at the end?

Local CSV files you can review and import into your existing budgeting workflow.

That includes categorized output plus Amazon clearing outputs.

Windows or Mac?

Windows is available now. macOS is coming soon.

Current target: within 2 weeks.

Amazon Automation Security

Amazon extraction runs through your own local browser session.

When you click Extract Amazon Transactions, UnboxedSpend opens a new browser window on your machine and works through that local session instead of asking you to type Amazon credentials into an UnboxedSpend form.

Usually already signed in

Amazon often keeps you signed in already, even after a restart, so extraction can begin without extra setup.

If Amazon asks, you log in there

If Amazon shows a login page, you sign in directly with Amazon in that browser window, then the automated extraction continues normally.

Credentials stay out of UnboxedSpend forms

Your Amazon username and password are not entered into an UnboxedSpend form.

This keeps the login step in Amazon's own browser flow while still letting UnboxedSpend automate the export once your local browser session is ready.

If Amazon changes the page structure and normal extraction breaks, UnboxedSpend first falls back to AI parsing. If that also fails, we develop an update fix and charge a small update fee for that repair.

Exports & Outputs

UnboxedSpend finishes by writing local CSV files you can review and reuse.

The finished outputs stay on your machine as CSV files, so you can inspect them locally and move them into your household budgeting workflow without waiting on a cloud export.

Categorized baseline

~/UnboxedSpend/output/Categorized/all_categorized_normalized.CSV

Your main categorized aggregate output, including the prized `ai_category` result that speeds up downstream budgeting work.

Amazon clearing output

~/UnboxedSpend/output/ClearingAccount/Amazon_Clearing_Account.CSV

Order-level Amazon clearing output built from the categorized baseline plus transaction-history reconciliation.

Itemized clearing output

~/UnboxedSpend/output/ClearingAccount/Itemized_Amazon_Clearing_Account.CSV

The deeper item-level clearing file for households that want product-by-product Amazon detail.

These outputs are designed to be reviewed locally and imported into the budgeting process you already use.

Transparent AI Cost Models

The UnboxedSpend Privacy Standard

Managed credits are the simplest default for most households. Bring your own API key stays available if you prefer direct provider billing and more control.

One-Click Link (OpenAI)

Best for: Speed and Simplicity

Sign in with your OpenAI account in seconds for a strong balance of privacy and convenience without any technical setup.

  • Link via Login

    Sign in with your OpenAI account in seconds. No technical keys or setup required.

  • Direct Pricing

    Pay OpenAI directly for exactly what you use. There are no added usage fees or markups from our side.

Direct API Link (BYOK)

Best for: API Key Holders and Azure Users

Connect your existing Azure OpenAI, OpenAI, or Google API keys directly and keep full control over the provider connection.

  • Full Key Control

    Connect your existing Azure OpenAI, OpenAI, or Google API keys directly.

  • Direct Pricing

    Use your existing credits or pay-as-you-go keys at your provider's standard rates. We do not charge additional fees for your AI usage.

Managed Gateway

Best for: Zero-Setup Convenience

Prefer a ready-to-use setup? Our Managed AI tier uses a secure Azure-hosted AI Gateway to build your rules for you without requiring separate AI accounts or API keys.

  • No Extra Accounts

    No need to manage external AI accounts or API keys.

  • Simple Refills

    Top up your rule-building credits directly from your UnboxedSpend dashboard as needed.

Example Household AI Spend

These are example budgeting ranges, not fixed quotes. Actual provider spend varies with model choice, file size, and how many categorization requests you run in a month.

Light month

Approx. Credits

10 to 25 credits

Approx. Dollars

About $1 to $3

Simplest Billing Path

Managed credits or direct provider billing

A smaller month with a few CSV imports or a short catch-up session.

Typical month

Approx. Credits

30 to 75 credits

Approx. Dollars

About $3 to $9

Simplest Billing Path

Managed credits are usually the easiest fit

A normal household reconciliation rhythm with regular categorization and occasional cleanup.

Heavy catch-up month

Approx. Credits

80 to 200 credits

Approx. Dollars

About $8 to $20

Simplest Billing Path

Use managed credits or your own API key for higher-volume months

A backlog-clearing month with more imports, retries, and extra categorization passes.