July 2026 Summaries

June was a significant month for the Apache Superset project, marked by substantial contributions and advancements across various areas. A total of 132 contributors merged 767 pull requests, highlighting key enhancements like the maturation of the extensions framework, which now includes a sanctioned method for AI chatbots to integrate with Superset via SIP-214. Security was another focus, with improvements in guest token management, password policies, encryption, and WebSocket layers. The project also saw an expansion in the MCP tool catalog with new dashboard automation tools and metrics, as well as enhancements in charts and visualizations, including cross-filter support and configurable options. Superset's community continued to grow, adding 376 GitHub stars, 57 new contributors, and 74 new members to the community Slack, with 68 individuals contributing their first pull requests. Additionally, internationalization efforts progressed with updated translations, and core infrastructure saw updates with Node and migration to Vitest for testing frameworks.

In the landscape of Business Intelligence (BI) tools, Preset stands out by offering a unique AI integration through its Model Context Protocol (MCP), which allows AI agents to not only read but also write back to data, enabling the creation of reusable virtual datasets. Unlike other BI vendors such as Tableau, Power BI, and Looker, which primarily offer read-only MCPs, Preset's MCP can create and save governed datasets that persist beyond individual AI sessions, enhancing collaboration and data continuity. Preset's MCP, built on the open-source SIP-187 specification, supports broad compatibility across AI clients and cloud environments, ensuring flexibility and avoiding vendor lock-in. The platform provides robust security measures, including OAuth 2.0, role-based access control, and full audit logging, making it suitable for enterprise-grade applications. This approach positions Preset not just as a tool with AI features but as a foundational component in AI-native workflows, emphasizing its role as both a destination for in-product experiences and a source for governed data layers that AI tools can query and build upon.