Skip to content
part of the groupApertiaDraivix

01Platform and tools · module

Data import and migration

Controlled import from CSV, Excel or API with preview, error logging and rollback

Core available, part in progressBook a consultation
  • Data source
  • Import definition
  • Import run
  • Import row error
15
capabilities
4
record types
5
automations
7
integrations and AI tools

02What the module does

Less routine. More value.

Data import provides a controlled way to bring data into the system from CSV, Excel, another record type or an API. First you see a preview without writing any data; then the import runs as a tracked job showing the counts of created, updated and failed rows, and the records it created can be rolled back. The same engine powers imports from the spreadsheet editor, AI-assisted imports and migrations from Shoptet and WooCommerce.

  • Data from files and APIsImport from CSV and Excel files, from another record type, or from an HTTP API—the interface of another application. Bring data into the system without manual retyping.
  • Columns mapped cleanly to fieldsMap source columns to target fields and add transformations and validation rules. Data enters the system in the exact format you expect.
  • Predictable conflict handlingFor conflicts, choose between create, update, create or update, or skip. You know in advance what will happen when a record already exists in the system.
  • No duplicate records createdBased on the fields you choose, the system recognizes when a record already exists and handles existing data according to the conflict mode you configured.
  • Preview first, write secondThe preview loads the first 10 rows, applies mappings and transformations, and shows the outcome without writing anything. Catch configuration mistakes before they touch your data.
  • Real-time import progressData is processed in batches (100 rows by default, up to 1000 rows) while the import displays live percentages and row counts. You never have to guess how the transfer is progressing.
  • Bad rows never halt the importAn invalid row is logged with its row number, original data, error code and a clear message while the import continues. You know exactly what needs fixing.
  • Clean and transform data on the way inBuilt-in functions format text, numbers and dates, look up values in reference tables and evaluate conditions. You don't have to clean inconsistent spreadsheets by hand beforehand.
  • Automatic completion alertsWhen finished, an import can send a notification or call a webhook to notify another system automatically. You don't need to watch and wait for it to finish.
  • Stop a running import anytimeCancel a running import and it will stop between batches. If you spot an issue with the source data, you don't have to wait for the entire run to finish.
  • Rollback after importA completed, failed or canceled import can be rolled back: the system deletes the records it created. Records that were updated by the import are not restored to their previous state.
  • Imports respect user permissionsEvery import runs under a specific user and their permissions: creating records requires create rights, and updating requires edit rights. It never grants anyone broader access.
  • Idempotent runs prevent double importsSubmitting the same request repeatedly returns the existing run instead of starting a new one. Accidentally sending a request twice will not duplicate your data.
  • Stalled runs never hang unnoticedThe system re-queues stalled pending imports and terminates runs whose processing has stopped responding. You always see the real status, and imports never hang silently.
  • Sealed data source credentialsCredentials for external data sources are sealed, and the interface only shows whether they are set. Anyone opening the settings cannot view the credentials themselves.

What is currently in progress

  • Database sources and JSON or XML files do not have a reader in this edition; you can import from CSV, Excel, another record type and APIs.
  • The module does not offer automatic scheduled recurring imports.
  • Rollback only deletes records that the import created; records that the import updated are not restored to their original state.
  • Migrations from Shoptet and WooCommerce cannot be rolled back.

03How it works

How it works step by step.

  1. How the import works

    First, you define the data source (a file or an API) and the import definition: target record type, column-to-field mapping, transformations, validation rules and conflict resolution mode: create, update, create or update, or skip. Based on the fields you select, the system detects whether a record already exists.

    Read in detail

    A preview loads the first 10 rows, applies the mapping and transformations, and shows the result without writing anything. Once started, the import processes data in batches (100 rows by default, up to 1000 rows) and continuously displays the percentage, current step, and counts of created, updated, skipped and failed rows.

    An invalid row does not stop the import. It is saved with its row number, original data, error code, clear error message and field name, so you know what to fix.

  2. Transformations and follow-up actions

    You can adjust values using built-in functions: uppercase and lowercase, whitespace trimming, numbers, dates, yes/no booleans, default values, splitting and joining text, string replacement, regular expressions, prefixes and suffixes, email and URL validation, lookup in reference tables, value mapping, conditional logic and combining fields.

    Read in detail

    Upon completion, the import can send a notification or call a webhook. You can also trigger workflows from import events—for example, to alert an administrator if an import fails.

  3. Security and traceability

    Every import is tied to a specific user and runs under their permissions: creating records requires create permissions, and updating requires edit permissions. An import triggered via an API token never exceeds the token's scope. Settings are frozen when the run starts, so editing the definition later does not affect a running import.

    Read in detail

    Submitting the same request repeatedly returns the existing run instead of starting a new one. Every minute, the system re-queues stalled pending imports and terminates runs whose processing has stopped responding. Orphaned files are deleted after 24 hours. Data source credentials are sealed, and the screen only indicates whether they are filled in.

  4. Cancellation and rollback

    You can cancel a running import, and it will stop cleanly between batches. A completed, failed or canceled import can be rolled back: the system deletes the records created by the import. Rollback requires permission to delete the target records and runs under the user who requests it.

04What runs automatically

Automations that save time.

  1. Pending imports are never forgotten

    Every minute, the system re-queues stalled pending imports. An import that failed to start will not sit stuck in the queue unnoticed.

  2. Timed-out runs are terminated

    Every minute, the system terminates runs whose processing has timed out. Your overview never shows an import pretending to run when it has actually stopped.

  3. Nightly cleanup of import files

    Every night, the system cleans up orphaned import files older than 24 hours. Storage does not fill up with files that no longer belong to any import.

  4. Running imports stay frozen

    When an import starts, its configuration is frozen. Editing the definition later does not affect a running import, so the result matches what you launched.

  5. Imports trigger follow-up workflows

    An import run emits events that you can hook workflows into—for instance, to notify an administrator automatically if an import fails.

External integrations

  • Data from external application APIsAn HTTP API—the interface of another application—can serve as an import source. The system validates the target address so requests cannot be sent to unauthorized destinations.
  • Migration from ShoptetThe Shoptet migration imports products, customers and orders from export feeds. You don't have to create them manually—just keep in mind that this migration cannot be rolled back.
  • WooCommerce e-shop in the systemThe import transfers your product catalog, customers and orders from WooCommerce. There is no manual entry required—just note that this migration cannot be rolled back.

What the AI agent handles

  • Hand the file to the assistant

    The assistant uploads an import file directly from the conversation. You can start preparing your data right in the chat without hunting for the right screen.

  • Check import status with one question

    Ask the assistant and it will check the status of an import run. You know how the import is progressing without looking it up in the log.

  • Rollback under your permissions

    The assistant can roll an import back, always acting under your permissions. It will never delete anything you wouldn't be allowed to delete yourself.

  • Open a spreadsheet editor from chat

    The assistant opens a bulk spreadsheet editor over your data. You can then edit large sets of records in a grid, just like in a spreadsheet app.

A person approves whenever money, a customer, or an exception is involved.

05Frequently asked questions

Can I test an import with a dry run first?

Yes. The preview loads the first 10 rows, applies the mapping and transformations, and shows the result without writing any data.

What happens if there is an invalid row in the file?

The import continues and saves the invalid row with its row number, data, error code and message. Once finished, you get a clear list of errors to fix.

Can an import be rolled back?

Yes, for standard imports. The system deletes the records created by the import; updated records are not restored, and migrations from Shoptet and WooCommerce cannot be rolled back.

Can Draivix import data from Excel?

Yes. The import tool supports both Excel and CSV files, data from another record type, and HTTP APIs.

Can an import bypass user permissions?

No. Every import runs under a specific user and their permissions, and an import via an API token never exceeds the token's scope.

06Related modules

Modules that work great together.

All modules in Platform and tools

Let’s start together

Tell us how your business works.

We will show you what a system that takes over part of that work could look like.

Book a consultation