These are the Treasure Data features and improvements released in August, 2025.
This feature enables users to extract nested fields from incoming JSON events and promote them to the top level via the Real-Time API, making those fields immediately available for segmentation, triggers, personalization, and reactor workflows.
Designed for data engineers and teams ingesting structured events, it eliminates the need for custom ETL, Lambda functions, or upstream flattening, especially when working with complex or third-party JSON event sources.
To learn more and access setup instructions, visit the Treasure Data API Documentation site.
This feature isolates SQL queries generated by the Large Language Model (LLM) into a dedicated Trino resource pool, ensuring they do not affect the customer's regular query concurrency.
Customers, especially Treasure Data administrators, can maintain consistent performance for their standard queries, with the LLM pool supporting contract-level concurrency.
To learn more, visit the Treasure Data Product Documentation site.
Treasure Data users can now generate enhanced dashboard-style outputs in the Foundry Workspace: similar to a screenshot or unified report view.
This improvement goes beyond simple text or single chart outputs, allowing agent users to organize and visualize data more effectively in comprehensive, visually rich layouts.
To learn more about this feature, please visit the Treasure Data Product Documentation site.
Claude Sonnet 4 is now live in the US region, delivering advanced AI capabilities for software development, including support for complex coding tasks and up to 64K output tokens.
Developers in the US can utilize Claude Sonnet 4 to accelerate new feature implementation with richer outputs and faster feedback.
Learn more by visiting the Treasure Data Product Documentation site.
This feature allows brands to deliver personalized website experiences in real time by making secure client-side API calls directly from the web browser, with robust token management and CORS support.
Data engineers and marketers can power use cases like loyalty-driven offers, dynamic product recommendations, and tailored messaging for e-commerce, all while ensuring sensitive user data is protected and never exposed in client-side responses.
Learn more by visiting the Treasure Data Product Documentation site.
This new feature enables users to download data from charts and graphs in Utilization Dashboards directly in CSV format.
Treasure Data customers can now independently acquire utilization data for deeper analysis or reporting, reducing the need for support requests.
For details and instructions, visit the Treasure Data Product Documentation site.
This feature enables advertisers and marketing teams to upload offline conversion event data (such as in-store sales, visits, or custom events) from Treasure Data directly to The Trade Desk via the CRM Data API.
Designed for Marketing Operations professionals, it provides an efficient way to attribute and analyze real-world conversions within digital ad campaigns, especially in verticals where purchases occur offline.
Learn more by visiting the Treasure Data Product Integrations Portal for more details.
This integration enables users to create new audiences and upload user data, as well as append data to existing audiences on NextRoll, directly from Treasure Data.
Designed for marketing operations, channel marketers, and data scientists who need to optimize audience management and campaign targeting within AdRoll/NextRoll platforms.
Learn more by visiting the Treasure Data Integrations Portal.
This feature allows marketers and data teams to programmatically ingest Meta Conversions API data quality metrics, such as event match quality, coverage, deduplication, and data freshness, directly into Treasure Data.
Marketing operations and analytics users benefit by monitoring, analyzing, and troubleshooting their Meta Conversions API integrations at scale, supporting improved campaign optimization and reliability across multiple Pixels and accounts.
Learn more by visiting the Treasure Data Integrations Portal.
The Google Drive Output Connector allows data engineers to export large query results from Treasure Data directly to their Google Drive in CSV format, supporting optional gzip compression and file splitting based on size or column count.
This feature is designed for users who need to share extensive reports or datasets within their organization, overcoming limitations found in Google Sheets and enhancing collaboration efficiencies.
For details and instructions, please visit the Treasure Data Integrations Portal.
This new connector enables users to export audience data, including custom attributes, from Treasure Data into Amplitude for identity management and profile enrichment.
Marketing, data science, and operations teams can now synchronize enriched audience profiles to Amplitude, supporting analytics, segmentation, and personalized experiences.
Learn more by visiting the Treasure Data Integrations Portal.