The Smart Import Engine

High-fidelity data migration with automated collision detection.

Step 0: The "Pluck" Logic

Importing data shouldn't mean cleaning up duplicates manually. Our engine performs In-Memory Collision Detection before a single row hits the database.

By analyzing the "Year-Month Buckets" of your file, the system fetches existing records and compares them against your upload. If an exact match for date, amount, and category is found, the duplicate is "plucked" out—leaving only fresh data to be saved.

Deduplication Rules

  • Transactions:Matches on Day, Amount, Category, and Memo.
  • Snapshots:Matches on Month, Account Name, and Balance.

Automated Sanitization

1. Currency Stripping

The engine automatically strips symbols ($), commas, and whitespace. "$ 1,200.00" is safely converted to a pure numeric 1200.00.

2. Text Truncation

To prevent database crashes, long memos or notes are gracefully trimmed to 100 characters while preserving the core detail.

3. Active State

All imported items are set to "Active" by default, ensuring they appear immediately in your current month's reporting.

Year-Month Bucketing

Transactions are processed based on their actual day to provide a granular history. Snapshot Items, however, are automatically aligned to the First of the Month (YYYY-MM-01). This creates a standardized "Snapshot" of your net worth that remains consistent across all historical reports.