Changelog

March 24, 2026 v1.5.0

Feature AI strictness modes — Choose between Strict, Moderate, and Flexible modes to control how aggressively the AI suggests value mappings. Strict only maps exact synonyms; Flexible handles creative supplier names like "midnight" → "black."
Feature Per-attribute AI toggle — Enable or disable AI suggestions on individual attributes. Keep AI active for color and material mapping while turning it off for SKU fields where you want manual control.
Improvement AI suggestion confidence scores now display as percentages alongside each proposed mapping, making it easier to review batch suggestions at a glance.
Fix Resolved an issue where AI suggestions would occasionally duplicate an existing value mapping entry.

February 18, 2026 v1.4.0

Feature Conditional transform rules — Apply transformations only when conditions are met. For example: "If category is Furniture, convert weight to kg" or "If brand is empty, set to supplier name."
Feature Regex replace — Use regular expressions in transform rules for advanced text manipulation. Strip HTML tags, reformat SKUs, extract numeric values from free-text fields.
Feature Live preview — See the output of your transform rules on real data before saving. The preview panel shows before/after for the first 25 rows so you can validate logic immediately.
Improvement Transform rules now execute in a user-defined order with drag-and-drop reordering.

February 4, 2026 v1.3.0

Feature Completeness scoring — Every processed feed now receives a completeness score (0–100%) based on how many required and recommended fields are populated across all rows.
Feature Missing field alerts — Automated alerts flag rows where required attributes are empty or contain placeholder values. Alerts are grouped by field so you can fix issues in bulk.
Improvement The feed detail page now shows a summary bar with total rows, mapped columns, unmapped columns, and new values requiring review.
Fix Fixed a bug where feeds with more than 50,000 rows could time out during the validation step.

January 21, 2026 v1.2.0

Feature XLSX export — Export normalized feeds as .xlsx files in addition to CSV and XML. Column types, number formatting, and sheet names are preserved.
Feature Split measurement columns — Optionally split measurement attributes into separate value and unit columns on export. "450 mm" becomes two columns: width_value (450) and width_unit (mm).
Improvement Export templates can now be saved and reused across feeds, so you only configure your output format once.
Fix CSV exports now correctly escape fields containing commas and newlines per RFC 4180.

January 7, 2026 v1.1.0

Feature Five access levels — Assign team members one of five roles: Owner, Admin, Editor, Reviewer, or Viewer. Each role has clearly scoped permissions for feeds, adapters, exports, and workspace settings.
Feature Workspace isolation — Data, adapters, and export configurations are fully isolated between workspaces. Team members can belong to multiple workspaces with different roles in each.
Improvement The settings page now includes a team management panel with invite links, role assignment, and an activity log.

December 10, 2025 v1.0.0

Feature Core normalization pipeline — Upload supplier feeds in CSV or XML format. FeedPrep parses, validates, and normalizes product data through your configured adapter rules.
Feature Column mapping — Map supplier column names to your canonical schema attributes. Mappings are saved per adapter and applied automatically on every future upload.
Feature Value mapping — Translate supplier-specific values to your standardized vocabulary. Supports select, multiselect, and boolean attribute types with AI-assisted suggestions.
Feature Unit conversion — Automatic conversion of measurements (weight, dimensions, volume) and formats (dates, decimals) to your preferred standards.
Feature CSV and XML export — Export normalized product data in CSV or XML format, ready for import into your webshop, PIM, or ERP.

See FeedPrep in Action