Release Summary 25.19 | Sep 18, 2025
The following key features and improvements, along with bug fixes, have been released in Algonomy CXP products in the release version 25.19.
Recommend
Compound Strategies Page Redesigned for Better Usability
The Compound Strategies interface has been rebuilt using Vue.js to improve usability, resolve pagination issues, and boost performance. The updated list view now includes a search bar, alphabetical sorting, column-based sorting, and clear action icons for edit and delete. Pagination has also been fixed.
The detail page retains the existing configuration flow, with enhanced interactivity. Users can now easily add, sort, and manage strategies with intuitive controls and real-time updates.
Jira: ENG-30639
Ensemble AI
Ensemble AI Default Prompt Now Pulled from Site Configuration
The Ensemble AI interface now loads the default prompt from the site configuration (ensemble ai llm prompt) for Free Form and Mixed Mode styles. Users can expand the prompt to view long messages, and internal employees may edit it at the style level if needed. Customers currently do not have edit access.
Jira: ENG-30963
Social Proof
Support for Badging in New Social Proof API
The Social Proof Experience Output API now supports badging, enabling server-side and app-based clients to access badge experiences alongside messaging. For any given product input, the API evaluates all associated badge criteria based on product attributes, identifies qualifying badges, applies priority rules, and limits display based on the configured count.
All matching badges that meet the criteria are returned in the API response, based on how the badging experience is set up. This allows for fully functional and configurable badge support within the new API.
Jira: ENG-27956
Engage and Social Proof
Context Builder Now Supports Price Constraints
You can now define price-based constraints in the Context Builder, enabling more tailored Social Proof messages based on a product’s price range. A new “Manage Constraints” option has been added to the Context Library for Brand, Category, Product, and Product Attribute conditions. This opens a popup where users can apply price filters using “More than,” “Less than,” or “Between” logic. This enhancement allows more precise targeting based on product pricing.
Jira: ENG-30823
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 25.19.
Jira # |
Module/Title |
Summary |
General Availability |
---|---|---|---|
Social Proof: Enable Optimization for Older Experiences in Social Proof
|
The Social Proof UI now supports enabling optimization at the variation level for older experiences and templates. A new site-level configuration—"enable social proof optimization for older experiences"—allows merchandisers to activate optimization without creating new experiences. Once enabled, the optimization toggle becomes available within legacy variations. Server-side integrations support this without changes, while client-side integrations may require template updates for full functionality. |
18-Sep-25 |
|
Engage: Content Test Drive Now Supports Multiple Image Attributes |
The Content Test Drive now correctly handles multiple attribute names specified in the “Content Image Attribute Names” site configuration. If a comma-separated list of image attribute names is provided, the system will use the first available attribute with a valid image to display in test results. This ensures images render reliably when multiple attributes are configured. |
18-Sep-25 |
|
Find: New Snapshot Type Control in SFI |
SFI now supports filtering of snapshot types using new configuration properties. You can specify which snapshot types to allow or ignore via the allowed-snapshot-types and ignored-snapshot-types settings under the find-indexer-streaming path in application.conf. This gives you fine-grained control over event processing for different snapshot types, such as excluding metastore events when needed. |
18-Sep-25 |
|
Find: Schema Patch Support Extended to Additional Item Types in SFI |
SFI now supports property definition patching for additional item types including suggestions, query vectors, and query tags, in addition to products and content. This enhancement allows scenarios such as no-code activation for related searches, where a new property can be added to an active metastore snapshot for a suggestion item type and be automatically reflected in the Consul schema. |
18-Sep-25 |
|
Find: Site-Level UER Weight Configuration Enabled |
UER (User Engagement Ranking) weights for purchases, views, and clicks can now be customized per site. This allows teams to fine-tune ranking behavior, such as prioritizing clicks for offer-heavy catalogs. Updates are managed via Consul and take effect when the Global Rank Publisher runs. |
18-Sep-25 |
|
Find: Configurable Snapshot Type Handling in SFI |
Support has been added to allow or ignore specific snapshot types in the Streaming Find Indexer (SFI). Two new config entries, allowed-snapshot-types and ignored-snapshot-types, let teams control which snapshot events get processed. This helps filter out unwanted snapshots like "metastore" for more efficient indexing. |
18-Sep-25 |
|
RR Server Now Routes Inference API Calls via Proxy |
RR Server has been updated to route SP Inference API calls through a proxy in the production environment, these calls continue to use the Marathon service configuration. This ensures consistent routing behavior across environments. |
18-Sep-25 |
|
Scoped Action Metrics Enhanced with Incremental Ingest Tracking |
The scoped action status response now includes a new metric incremental_touched to track items ingested from external channels during an active ProductMatch scoped action. This separates them from items directly updated (touched) or untouched (untouched), offering more accurate reporting. |
18-Sep-25 |
|
Ecommerce Sales Assistant Chatbot: Full Site-Level Configuration Support |
The Sales Assistant chatbot can now be fully configured for new clients directly from the LLM Configuration interface. eCommerce managers can specify tools, descriptions, gender options, agent prompts, and sample query-style mappings to simulate real product responses. These configurations are stored site-wise via the portal API and consumed during chatbot processing, allowing complete preview and alignment with brand requirements without backend intervention. |
18-Sep-25 |
|
Ecommerce Sales Assistant Chatbot: Expanded Configuration Options |
The Ecommerce Sales Assistant chatbot setup page now includes flexible configuration fields for API keys, site details, product filters (e.g. gender, budget), tooltips, agent prompts, and query-style examples. These updates allow eCommerce managers to fully tailor chatbot behavior and output without backend support, streamlining deployment and customization. |
18-Sep-25 |
|
ENG-30809, ENG-30810, ENG-30811, ENG-30812, ENG-30813, ENG-30814 |
Dashboard Chatbot: Expanded Product Documentation Coverage |
The chatbot now includes documentation from additional modules beyond Recommend, Strategies Library, Integration, and Personalization Cloud. Newly added modules include Find, Discover, Streaming Catalog, Experience Insights, Testing (MVT), Engage, Guided Selling, Catalog Enrichment, Social Proof, and Ensemble AI. Note: This feature is currently available to employees only and will be rolled out to customers in a future release. |
18-Sep-25 |
Bug and Support Fixes
The following issues have been fixed in the release version 25.19.
Jira # |
Module/Title |
Summary |
General Availability |
---|---|---|---|
Date Format Mismatch for FTP Document Read |
Resolved an issue where the application could not read documents from the FTP server due to a date format mismatch. The date format has now been aligned to ensure weekly automated document processing works as expected. |
18-Sep-25 |
|
Find: Personalization Exception in Solr 9 |
An exception previously fixed for standard requests was missed when personalization was applied, causing failures during Solr 9 search queries. This issue has now been resolved. Personalized search requests are executing successfully, supporting Solr 9 compatibility. |
18-Sep-25 |
|
Data Engineering , Social Proof: Product Separator Handling in Social Proof API |
We have fixed an issue in the spMessages API where it was returning an empty array when multiple product IDs were passed using the pipe (|) separator, particularly for category and cart page placements. The API now correctly returns all active campaigns and messages for the listed products without errors. |
18-Sep-25 |
|
Data Engineering , Social Proof: Fix for Location Selector and Importance-Based Sorting in Badging API |
We have fixed an issue where the location Selector for item pages was incorrectly showing extra attributes. It now displays only the correct element type as configured. Additionally, sorting of badges and messages is now unified based on the importance value, ensuring a consistent display order. |
18-Sep-25 |
|
Ensemble AI: Missing and Duplicate Seed Products in Ensemble AI |
We have resolved an issue in Ensemble AI where the API did not return the complete list of seed products and occasionally returned duplicates. The problem stemmed from a newline character in the backend response, which caused the last product in the list to be skipped. Additionally, when the API was called without a style ID, duplicate entries were appearing. Both issues have now been fixed, ensuring the seed product list is accurate and complete. |
18-Sep-25 |