January 2026 Summaries

Data tool fragmentation often leads to redundant efforts in defining business logic across various platforms, a challenge that Preset aims to address through its involvement in the Snowflake-led Open Semantic Interchange (OSI) initiative. This initiative seeks to develop a vendor-agnostic specification for sharing semantic models across different tools, enhancing interoperability and reducing redundancy. Concurrently, Preset is advancing Apache Superset's capacity to handle semantic layers more effectively by proposing first-class support through SIP-182, allowing for more seamless integration and reducing the awkwardness of current pseudo-database implementations. The OSI initiative complements this by standardizing how semantic definitions are exchanged between tools, easing migrations, enhancing AI application context, and facilitating better exploration workflows. Preset's contributions, including integrating Superset with semantic layers like dbt MetricFlow and Cube, inform their work on creating open adapters and advocating for improved exploration capabilities. As these developments unfold, community feedback and participation will be crucial in shaping the future of semantic interoperability and user experience in data exploration.

Semantic layers, previously sidelined in data analytics, are regaining popularity due to their potential to facilitate trustworthy and structured AI interactions and improve self-service analytics. Historically hindered by issues such as tool lock-in and the bottleneck of data teams, semantic layers are now shifting from BI tools to the transform layer, allowing for version-controlled, auditable, and portable code. The resurgence is driven by the increasing need for AI agents to have structured data access, as exemplified by new solutions like dbt's semantic layer and open standards like SDF. Companies like Preset are embracing this shift by supporting various semantic layers in their tools, aiming to overcome market inertia and empower organizations to explore and invest in semantic layers without fear of technological dead ends. This approach allows businesses to balance curated, governed data experiences with the flexibility of self-service analytics, facilitating a broad range of use cases from mature, battle-tested metrics to more exploratory data scenarios.

December 2025 was a busy month for Apache Superset, marked by significant contributions from 42 contributors, including 19 first-timers, resulting in 266 pull requests. Key highlights included advancements in the extensions system, security enhancements, and improvements across SQL Lab, dashboards, and chart functionalities. The extensions system became more robust, featuring new community extensions like the SQL Flow Visualizer and SQL Lab Export to Google Sheets, among others. Security updates included configurable hash algorithms for FedRAMP compliance and fixes for Row-Level Security vulnerabilities. Dashboard improvements focused on visual and interaction enhancements, while SQL Lab saw updates in functionality and permission fixes. Chart enhancements included updates to table charts and heatmap y-axis sorting. Efforts were also made in developer experience, such as TypeScript migration and Docker development enhancements. Additionally, updates were made to documentation, database support, configuration enhancements, and UI/UX improvements. The community was encouraged to participate further with resources like the Contributing Guide and Community Slack, and the month concluded with a look forward to the developments expected in 2026.