# May 2026 Release Note These are the Treasure AI features and improvements released in May, 2026. ## Engage Studio ### New Feature: Secure, Non-PII Engagement Event Logging *Posted on: May 21, 2026* - Engage Studio now supports non-PII identifier logging for email engagement events — including delivery, bounce, error, and unsubscribe events — so you can correlate events back to customer records without storing raw email addresses at rest. - Designed for enterprise customers with strict internal security policies that prohibit persisting raw PII in analytics or operational tables, this feature lets you choose which non-PII identifiers (such as a hashed customer ID) are logged, preserving analytics, unsubscribe management, and operational traceability while keeping sensitive identifiers out of downstream tables. - Learn more by visiting the [Email Delivery Events Table](/products/marketing-cloud/engage-studio/channels/email/email-delivery-events-table) page. ### New Feature: Dedicated Status Page Components for Engage Studio APIs *Posted on: May 26, 2026* - The Treasure AI [Status Page](https://status.treasure.ai/) now includes two dedicated components — **Engage Studio - Campaign API** and **Engage Studio - Delivery API** — providing independent operational visibility into Engage Studio's campaign management and message delivery operations. - Designed for marketers and IT administrators who run customer engagement programs on Engage Studio, this enhancement makes it easier to isolate, communicate, and subscribe to incidents and maintenance windows scoped specifically to Engage Studio rather than tracking them through general platform components. - Subscribe to either component on the Status Page to receive notifications about uptime, incidents, and scheduled maintenance for the Engage Studio APIs you depend on. ## Integration Hub ### New Connector: Shopify Webhook Streaming Ingress *Posted on: May 20, 2026* - The Shopify Webhook Streaming Ingress connector is a new managed streaming source that ingests native Shopify webhook events directly into Treasure AI in near real time, with built-in HMAC-SHA256 signature verification and automatic flattening of Shopify's deeply nested JSON payloads. - Designed for ecommerce teams using Shopify alongside Treasure AI Real-Time 2.0, this connector eliminates the need to build and maintain custom middleware (such as AWS EventBridge bridges). It supports order, customer, checkout, refund, cart, inventory, product, and collection topics, enabling low-latency use cases like cart abandonment recovery, low-inventory alerts, and next-best-offer triggers. - Learn more by visiting the [Shopify Streaming Import Integration](/int/shopify-streaming-import-integration) page. ### Enhancement to Google Cloud Storage Export Connector: Workload Identity Federation Support *Posted on: May 21, 2026* - The Google Cloud Storage Export connector now supports Workload Identity Federation as an authentication method, allowing Treasure AI to write to GCS buckets without requiring long-lived service account keys. - Designed for IT administrators and data engineers who export data to Google Cloud Storage, this enhancement strengthens security posture by replacing static credentials with short-lived, federated tokens issued through a Workload Identity Pool — reducing key-rotation overhead and exposure risk for organizations with strict cloud security policies. - To use Workload Identity Federation, configure your bucket with the **Uniform** access control model and create a Workload Identity Pool in Google Cloud IAM, then select the Workload Identity Federation authentication mode when configuring the connector in Treasure AI. - Learn more by visiting the [Google Cloud Storage Export Integration v2](/int/google-cloud-storage-export-v2-integration) page. ### Enhancement to Salesforce Marketing Cloud Export Connector: Row-Based Data Retention *Posted on: May 21, 2026* - The Salesforce Marketing Cloud (SFMC) Export connector now supports row-based data retention when creating new Data Extensions, letting you specify the number of days each row should be retained — independently of the Data Extension's default ~2-year retention policy. - Designed for marketers and data engineers exporting customer data to SFMC, this enhancement enables per-activation retention control (for example, 30 days) so personalization and engagement use cases can comply with data minimization requirements without manual cleanup of stale rows. - Enable row-based retention by setting `enable_row_based_retention=true` and `retention_period_days` to the desired number of days when configuring an export to a new Data Extension. - Learn more by visiting the [Salesforce Marketing Cloud V2 Export Integration](/int/salesforce-marketing-cloud-v2-export-integration) page. ## Customer Data Platform ### Enhancement to Real-Time Configuration: Launch Stage Progress Indicators *Posted on: May 18, 2026* - Real-Time parent segment configurations now display progress banners that indicate the current launch stage after you start or update a configuration, replacing the previous lack of visibility into configuration deployment status. - Designed for marketers and data engineers who manage Real-Time parent segments, this enhancement provides clear feedback through two stages: a Database Update stage (typically a few minutes) and a Reactor Upload stage (up to several hours depending on data volume), so you know exactly where your configuration stands. - While the configuration is launching, existing configurations continue to serve live traffic until the update completes, and a "View progress" link may appear to let you monitor the associated Data Workbench workflow. - Learn more by visiting the [Setting Up Parent Segments for Real-Time](/products/customer-data-platform/real-time/creating-a-real-time-configuration#configuration-launch-stages) page. ### New Feature: AI Skills for Real-Time Debugging *Posted on: May 18, 2026* - Treasure AI now provides six AI-powered debugging skills — `activations`, `identity`, `identify-top-key-values`, `id-graph-canonical-id-size`, `id-graph-ids-to-canonical-id`, and `rt-journey-monitor` — that generate diagnostic queries to help you quickly identify root causes of common Real-Time issues. - Designed for data engineers and marketers managing Real-Time configurations, these skills provide guided workflows for investigating activation failures, identity stitching problems, and journey execution issues without writing queries from scratch. - Install the `td-skills` package through Treasure Code to get started, then provide your parent segment ID along with a specific symptom to receive targeted diagnostic output. - Learn more by visiting the [AI Skills for Real-Time Debugging](/products/customer-data-platform/real-time/real-time-ai-debugging-skills) page. ### New Feature: Per-Profile ID Limit for Real-Time ID Stitching *Posted on: May 20, 2026* - Real-Time Stitching now enforces a limit of 200 unique IDs per profile, protecting both data quality and system performance by preventing profiles from accumulating an excessive number of stitched identities — a condition known as "over-stitching." - Designed for data engineers and marketers who configure Real-Time ID stitching keys, this safeguard ensures that profiles built from insufficiently unique key values (such as a generic region ID) do not grow unbounded and degrade system reliability. - When a profile reaches the 200-ID limit, the system automatically accepts new IDs and ejects the oldest ones first — bulk data is never deleted, but the association to the profile inside the Real-Time system is purged and can be rebuilt from new events or by re-running your ID graph workflow. - Learn more by visiting the [Real-Time ID Stitching Overview](/products/customer-data-platform/real-time/real-time-id-stitching-overview#per-profile-id-limit) page. ### Enhancement to Data Workbench: Hive DELETE Statement Support *Posted on: May 21, 2026* - The Hive query engine in Data Workbench now supports the `DELETE FROM ... WHERE` statement, which was previously available only in Trino — giving you a single, consistent way to remove rows from a table regardless of the engine you use. - Designed for data engineers who run large-scale data maintenance, GDPR/CCPA right-to-erasure workflows, or table cleanup jobs, this enhancement enables Hive-based DELETE operations on tables that may exceed Trino's memory limits, complementing existing data expiration and Trino DELETE workflows. - A `WHERE` clause is required, and DELETE statements cannot be issued against data stored in a Data Tank. - Learn more by visiting the [DELETE Statement Syntax](/products/customer-data-platform/data-workbench/queries/sql-reference/delete_statement) page. ### New Feature: No Re-Entry Option for Real-Time Journeys *Posted on: May 21, 2026* - Real-Time journeys now support a "Never allow re-entry" option that ensures profiles can only enter a journey once — after a profile exits, whether by meeting the goal, exit criteria, or completing the journey, the profile cannot re-enter even if it qualifies again. - Learn more by visiting the [Creating a Real-Time Journey](/products/customer-data-platform/journey-orchestration/realtime/creating-a-real-time-journey) page. ## Treasure Workflow ### New Feature: Workflow Trigger *Posted on: May 29, 2026* - Use Workflow Trigger to automatically start a workflow attempt when an upstream workflow's attempt succeeds. - Designed for data engineers and developers managing chains of dependent workflows (such as ETL pipelines with multiple downstream consumers), Workflow Trigger lets each downstream workflow declare its own dependency, removing the upstream-side maintenance of `call>`/`require>` and the timing guesswork of fixed-time schedules. - Learn more by visiting the [Workflow Trigger](/products/customer-data-platform/data-workbench/workflows/workflow-trigger) page. ## Control Panel ### New Feature: API Key Expiration Policy (GA) *Posted on: May 27, 2026* - API key expiration policies are now generally available, enabling administrators to enforce automatic API key rotation by configuring expiration lifespans at both the system-wide and individual user levels. - Designed for IT administrators and security teams managing programmatic access to Treasure AI, this feature reduces the risk of compromised long-lived credentials by automatically expiring API keys after a defined period (1–365 days), requiring users to create replacement keys periodically. - Expiration policies apply to both Master and Write-only API keys, with user-level settings overriding system-level defaults. Existing keys created before a policy is enabled retain their original "Never Expire" behavior until manually replaced. - Configure expiration policies through Control Panel > Security > API Keys, and set up automated expiration reminders to notify users before their keys expire. - Learn more by visiting the [API Key Expiration Policy](/products/control-panel/security/api-key/api-key-expiration-policy) page.