Release version 7.5.9
Release version 7.5.9 includes new features, improvements, and bug fixes.
- Feature updates
- Upcoming Product Enhancements
Skypoint AI
Think & Search Agent Integration: Skypoint AI Copilot now includes Search & Think feature, where Search enables real-time web searches, fact-checking, and market research for accurate, up-to-date responses, while Think provides provides better reasoning, action for greater AI accuracy, relevance, and efficiency.
Multiple SQL Agent Support: Skypoint AI Copilot now supports multiple SQL agents, enhancing automation, and AI-driven interactions while enabling users to integrate and manage multiple agents within a single Copilot instance, improving workflow efficiency and scalability.
Backup Configuration for AI Models: Skypoint AI has implemented a backup configuration system to ensure uninterrupted model availability in case of an outage, enabling automatic failover support for all AI models, including Gemini, Llama, and ChatGroq, enhancing system resilience and reliability.
Platform
Whitelist Email Domain on Skypoint AI Studio: Admins will soon be able to manage whitelisted email domains directly from the Skypoint AI Studio UI, eliminating the need for backend updates. This enhancement simplifies domain management, making it easier to configure automatic sign-in for users with SSO access.
Optimized Common Data Model: We have optimized the Common Data Model (CDM) by removing default extra columns from Audience, Profile, Metrics, and Timelines, ensuring only relevant source data is included, thereby improving data integrity, storage efficiency, and query performance.
CMS SNF Connector: The CMS SNF Connector leverages Azure Durable Functions to optimize CMS data imports into Azure Data Lake Storage, ensuring sequential execution, automated retries, and scalable processing for seamless performance.
Skypoint AI
- Library Upgrade: We are upgrading Python from version 3.10 to 3.13 to enhance performance, security, and compatibility while ensuring better efficiency, long-term support, and improved stability across applications.
Platform
Databases Explorer: We are upgrading the Database Explorer in Skypoint AI Studio to offer a more intuitive and comprehensive view of schemas and tables under Unity Catalog, improving data visibility, organization, and accessibility within the Lakehouse for a seamless and more efficient user experience.
Deprecation of Legacy Features: We are retiring the Data Map and Data Processing Activities features from Empower, as well as the Custom Models feature from Predict. This update streamlines the platform by focusing on core capabilities and optimizing user experience.
Streamlined access to Skypoint AI Copilot: We’re introducing a new role, Copilot Viewer, designed to provide users with access exclusively to Skypoint AI Copilot. This role will ensure precise access control, offering a streamlined experience tailored to specific needs.