Release Summary - Jun 04, 2024 (24.11)
The following key features and improvements, along with bug fixes, have been released in Algonomy CXP products in the release version 24.11 during May 17, 2024 - Jun 04, 2024.
Enterprise Dashboard, Science
Ensemble AI (Composite Outfit): Apply User Affinity to Re-Rank Outfits
Ensemble AI (Composite Outfit) now supports re-ranking outfits based on user affinity scores, enabling personalized recommendations for shoppers.
Using the existing User Affinity Configuration, the system can now re-rank outfits for each shopper according to their affinity scores. Merchandisers can configure Ensemble AI to point to a specific User Affinity Configuration, which is used to total the scores and re-rank the outfits. An API has been provided to set the User Affinity Configuration on the Ensemble AI list page.
If no User Affinity Configuration is set, the outfits will not be re-ranked by user affinity, and this configuration is considered valid. The User Affinity Configuration applies to all styles created within Ensemble AI.
Jiara: ENG-27628
Ensemble AI (Composite Outfit) - Update API with More Product Information
The Ensemble AI (Composite Outfit) API has been enhanced to provide additional product details, enabling merchandisers to use these details as part of active content.
The updated API now includes product details such as the product image URL, product name, brand name, product URL (linking to the item page on the client's website), and pricing information (including both sale price and list price). Additionally, the sort order of products within the composite outfits has been corrected to ensure consistency across ensembles, preventing issues where the products appear in different orders.
Jira: ENG-27977
Ensemble AI (Composite Outfit): Update APIs to Consider User IDs Based on User Affinity
Ensemble AI (Composite Outfit) APIs now support the inclusion of user IDs for personalization based on user affinity. This enhancement allows merchandisers to pass user IDs as part of the R3_COMMON or API, enabling tailored outfit recommendations for shoppers.
User IDs are now included in the composite outfit API to personalize the outfits according to the shopper's user affinity configurations. This update ensures that the rrserver API for the composite outfit (Ensemble AI) accepts necessary input parameters, including the external user ID (userId), at the template level.
Jira: ENG-28124
Ensemble AI (Composite Outfit): Attributes Auto-Suggestion for Filters
The Ensemble AI (Composite Outfit) feature now supports autosuggestions for product attributes when defining a style at the part level. As merchandisers type in the attribute field, a list of up to 20 relevant attribute names and values is displayed, making it easier to find and select the correct attributes without memorizing exact names. This auto-suggestion feature improves efficiency and reduces errors.
The same functionality has been applied to the brand autocomplete/dropdown, allowing for efficient brand selection.
Jira: ENG-28146
Data Engineering
Social Proof: Language-Based Personalization and Language Parameter Support
Social Proof now supports language-based personalization for messages, both through the UI and server-side integration. This enhancement allows users to view messages in their locale based on their language settings, providing a more tailored and relevant user experience.
Key Features:
-
Multi- Language Configuration: Merchandisers can configure social proof messages for multiple languages through a single template. For each language messages can be specified in the appropriate locale.
-
Localized Text Entry: Users can enter text for each specified language, ensuring messages are accurately localized. This feature is supported on the UI side, allowing for easy setup and management of localized messages.
-
Dynamic Messaging Display: Social proof messages dynamically display the appropriate text based on language parameters passed in the R3_COMMON configuration. This supports language (lang) parameters. If the language parameter is passed directly, the corresponding text is displayed.
-
Language Parameter in API: The Social Proof Output Response API now accepts a language parameter (lang), enabling the display of messages in the specified language. This allows for seamless integration with applications, ensuring that users receive messages in their preferred language through server-side interactions.
Social Proof - Update New SP API to Support Category Page for Multiple Message Types
The new Social Proof API has been enhanced to support multiple message types for category and list pages. This update allows Digital Optimization Managers to display the highest priority social proof message among multiple metrics, ensuring that a relevant message is shown even if one metric does not meet the threshold criteria.
When the "Enable Multiple Messages" option is disabled, the API will return only the highest priority message that meets the threshold. If the "Enable Multiple Messages" option is enabled, the API will return all messages that meet the threshold, ordered by priority.
Additionally, the API respects the show/hide option for each message type, ensuring that only the selected messages are included in the response.
Jira: ENG-27152
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 24.11 during May 17, 2024 - Jun 04, 2024.
|
Jira # |
Module/Title |
Summary |
General Availability |
|
|
Find: Run Query Understanding in Stages |
Query understanding now operates in four stages to improve result accuracy: QUERYTAG_OPERATION: Executes the initial query tag operation. QUERYTAG_OPERATION_WITHOUT_FILTER: Removes query tag filters if stage 1 yields no results. REFORMULATE_QUERYTAG_OPERATION: Reformulates the query if stages 1 and 2 fail. POST_QUERYTAG_OPERATION: Changes the query if zero results persist after all previous stages. |
04-jun-24 |
|
|
Science: Add UserId Support to the Dashboard Ensemble AI (Composite Outfit) API |
The dashboard Ensemble AI (Composite Outfit) API now supports the userId parameter, allowing for the inclusion of user IDs to enhance user affinity. This update ensures that user-specific data can be passed to the API for personalized experiences. |
04-jun-24 |
|
|
Recommend: Update Avro Schemas to Store External Product IDs for Glass Views |
The avro schemas have been updated to store external product IDs for glass views. Previously, UPS stored glass views using internal product IDs, while other view events used external IDs. |
04-jun-24 |
|
|
Recommend: Configurable Strategies -Respect Site Config for Previously Purchased Products |
Configurable Strategies have been updated to respect the "don't recommend previously purchased products" Site Config. Previously, this setting was ignored, impacting strategies like the new Ingredient Cross-sell.
Now, for sites where this Site Config is disabled, Configurable Strategies will not filter out previously purchased products. |
04-jun-24 |
|
|
Recommend, Recs+Email: Customize "Enable Request-Based Response Domain" to Skip Mail API Requests |
The "enable request-based response domain" feature has been updated to exclude MailView API requests. This change prevents email click URLs from including internal IP addresses, ensuring they remain functional.
For most API calls, click URLs will be generated based on the domain of the incoming request. However, email recommendation API calls will now generate click URLs based on the base rrserver configuration (e.g., recs.richrelevance.com for production). |
04-jun-24 |
|
|
Streaming - Recommend: Store Ancestors and Direct Children Data for Categories |
The Streaming Recommend feature now stores category_ancestors and category_direct_children data for categories, ensuring accurate recommendations based on complete category hierarchy data.
Parent categories must be ingested before child categories to maintain the correct hierarchy. |
04-jun-24 |
|
|
Update rr avro Version to Support Saving External ID for Mail Views |
The rr avro version has been updated to support saving external IDs for Mail Views. |
04-jun-24 |
|
|
Store Product external ID instead of internal ID for Email Views in UPS |
The system has been updated to store product external IDs instead of internal IDs for Email Views in UPS. |
04-jun-24 |
Bug and Support Fixes
The following issues have been fixed in the release version 24.11 during May 17, 2024 - Jun 04, 2024.
|
Jira# |
Title |
Summary |
General Availability |
|
|
Recommend: Unable to Save Browse Configurations in Context |
We have resolved an issue that prevented the saving of context in Browse Configuration. Users were encountering errors when trying to add multiple categories, causing the rule not to be saved. This issue was inconsistent, occurring with both single and multiple category additions. |
04-jun-24 |
|
|
Composite Outfit, Ensemble AI, Science: Ensemble AI - Not Generating Outfits for The Collective (2004) |
We have fixed an issue where Ensemble AI was not generating outfits for the specified style on The Collective site for a client. Despite the presence of products, no outfits were generated, even when using complementary colors. |
04-jun-24 |
|
|
Enterprise Dashboard: UI Recs Test Drive - Fixes |
We have fixed several issues in the UI Recs Test Drive to improve functionality and user experience.
Fixes Include: The rid parameter now correctly passes the region ID instead of the region name. The Placement dropdown now includes only placements where "hasRecs": true. The channel field has been renamed to API Client Key and is now a dropdown showing options from the /channels API, excluding default channels. In the rules tab, the 'Management UI' column header has been changed to 'Type' and strategy rules are displayed as "Strategy". When there are no recommendations, the request and an appropriate error message are displayed. If the force strategy option is used and the strategy returned is not the same as the forced strategy, a message is added stating, "The strategy that was selected was unable to play. There must be a strategy message for that strategy under the selected page type." |
04-jun-24 |
|
|
Data Engineering, Social Message: Social Proof Output Response API - Not getting appropriate response when Interval is not given in minutes |
We have fixed an issue with the Social Proof Output Response API where no response was generated if the interval was not provided in minutes. The API now correctly converts the interval as appropriate based on the input format. |
04-jun-24 |
|
|
Enterprise Dashboard: Unable to Create Segment with User Attribute Not in Auto-Suggest |
We have fixed an issue that prevented creating a segment using a User Attribute name not appearing in the auto-suggest. Users can now add a User Attribute name even if it is not suggested by the auto complete. |
04-jun-24 |
|
|
Enterprise Dashboard: Remove Duplicate Guided Selling Template in Dynamic Experiences |
We have fixed an issue where two Guided Selling Q&A templates were appearing in Dynamic Experiences. Now, only one Guided Selling Q&A template is displayed. |
04-jun-24 |
|
|
Enterprise Dashboard: UI Permissions Issue for Non-Employees in Recs Test Drive |
We have resolved an issue where non-employees were unable to use Recs Test Drive under certain conditions, such as when using a placement requiring a product ID or on sites using regions. The product ID and regions input fields were not appearing due to a missing requirejs library for non-employee portal users. |
04-jun-24 |
|
|
Science: ReplenishmentMR Job Continuously Failing Due to Jar Dependency Issues |
We have resolved an issue where the ReplenishmentMR Job was continuously failing in both production and QA environments due to jar dependency issues.
The issue has been fixed and the job ran successfully with patch 1.82.1, generating the respective files. |
04-jun-24 |
|
|
Data Engineering, Social Message: Social Proof Output Response API - Inconsistent Variable Names Issue Fixed |
We have resolved an issue with the Social Proof Output Response API where responses were not being generated due to inconsistent variable names. The API now correctly processes both singular and plural variable names, ensuring valid experiences are handled appropriately.
Previously, the API only supported plural variable names like "add_to_carts" and "purchases," while ignoring singular names such as "add_to_cart" and "purchase." This inconsistency has been fixed now. |
04-Jun-24 |
|
|
Data Engineering, Social Message: Social Proof Output Response API - Highest Priority Message Not Showing |
We have fixed an issue where the highest priority social proof message was not being displayed correctly. The API now properly prioritizes and shows messages based on the defined order, ensuring that the highest priority message is displayed.
Prioritization of Messages: Purchases last hour, Purchases today, ATC today, Views now, Views today. |
04-Jun-24 |
|
|
Enterprise Dashboard, Social Message: Social Proof - Category/Cart Page API Experiences Not Working |
We have fixed an issue where existing social proof experiences for category, cart, and list pages were not loading for some clients. |
04-Jun-24 |