Bulk Import

When you are setting up a new organization, migrating from another system, or onboarding a batch of new records, bulk import gets the data in efficiently. AI-native ERP provides a universal import system that handles all master data types through a single interface, with automatic duplicate detection, flexible entity resolution, and per-record error isolation.

Supported Data Types#

The import system handles all major master data categories. You can import customers, vendors, employees, GL accounts, items, dimension types, dimension values, and fixed assets. Each data type has its own set of required and optional fields, duplicate detection logic, and validation rules.

For chart of accounts, you can import your entire GL structure at once, including parent-child relationships for hierarchical accounts. Dimension values can be imported with hierarchical structures using nested tree formats. Fixed assets import with full depreciation configuration, including category assignment and useful life parameters.

Migration from External Systems#

When migrating from QuickBooks, Xero, or other accounting software, external ID support lets you preserve your existing record identifiers. You can include an external_id on each imported record, which the system stores and uses for future cross-references. This makes it easy to reconcile imported data against your source system and to import related records (like transactions that reference vendors) using the external identifiers.

Duplicate Detection#

Each data type has defined duplicate detection keys. Vendors are matched by tax ID first, then by name. Customers follow the same pattern. Employees are matched by work email. Accounts are matched by account number. Items are matched by item number. When a duplicate is detected, you control the behavior: skip the record, update the existing record with new data, or flag it as an error.

Validation and Progress#

For any import, you can run in validate-only mode first. This performs all the same checks -- required field validation, reference resolution, duplicate detection, business rule enforcement -- without creating any records. You review the validation results, fix issues in your source data, and then run the import for real.

Small imports (50 records or fewer) process synchronously and return results immediately. Larger imports process asynchronously, and you can check progress at any time. The progress report shows how many records have been processed, imported, updated, skipped, and errored, with specific error details for any failed records.

Reference Resolution#

You do not need to look up internal IDs before importing. The system resolves references by name, code, or external ID. If your vendor import includes a payment_terms_code of "NET30", the system finds the matching payment terms record automatically. If a referenced entity does not exist, the record is flagged with a clear error message rather than failing silently.