April 2025 Summaries

The text discusses the challenges of modern organizations with complex data stacks, leading to "data chaos" where conflicting metrics, redundant logic, and tool-specific workarounds slow down decisions and undermine trust. This chaos manifests in various ways, including inconsistent definitions of key metrics, redundant data modeling efforts, manual data exports, slow queries, and disjointed governance policies. To resolve this chaos, a universal semantic layer is needed to standardize logic, govern access, and accelerate performance across all analytics and AI tools. Cube Cloud introduces such a layer, allowing centralization of business logic, governance of access, and optimization of performance. This clarity enables decision-makers to trust the data, simplifies auditing, and improves user experience for executives, data teams, business users, and application development teams. With the increasing adoption of AI, real-time analytics, and embedded data experiences, enterprises must deliver consistency and confidence in their data systems to win.

We're announcing Query History export, a new Cube Cloud feature that lets you export Query History data to your existing monitoring and observability tools. With this feature, you can create custom reports on query performance, alert when queries fail or bypass configured pre-aggregations, analyze your semantic layer usage by different security contexts, archive and retain your Query History data indefinitely, and unlock insights into your Cube Cloud environment. This new feature is available on the M tier of Monitoring Integrations and provides a valuable tool for teams running a semantic layer at scale to troubleshoot long-running queries, optimize pre-aggregation usage, and build internal usage-based reporting.

AI governance refers to the framework of policies, processes, and controls that guide the responsible use of AI technologies. It ensures AI operates within acceptable boundaries, ethically, legally, and strategically. However, AI governance adds layers of complexity compared to traditional data governance, including model behavior and interpretability, prompt input and response output control, bias mitigation in generated content, versioning and change management of semantic and model logic, auditability of AI-driven decisions, security and compliance for data used in AI training or querying. The scope is broader and the consequences of failure more public than traditional data governance. AI systems need to interpret and apply governance consistently in real time, which can be a challenge for organizations. A universal semantic layer like Cube Cloud's is essential to enable AI governance by centralizing business logic, ensuring consistency, and enabling AI to generate responses that are trusted, auditable, and aligned with enterprise standards.

The text discusses the "confidence gap" in modern organizations, where teams struggle with inconsistent and untrustworthy data. This gap is caused by disconnected systems, duplicated logic, and poor tooling, leading to delayed decisions, redundant work, shadow workarounds, and eroded trust. To close this gap, a universal semantic layer is needed to standardize how data is defined, accessed, and interpreted. Cube Cloud is proposed as a solution that eliminates these root causes by centralizing metric definitions, providing consistent access to data, improving performance, enforcing security and governance policies, and rebuilding trust in the data.

Our mission at Cube is to use consistency, context, and trust to enable the next generation of data experiences. We are innovating daily for our clients and the community of data engineers and app development teams who use Cube to unify, govern, optimize, and integrate across the data stack and an increasing number of data consumers, such as AI, BI, spreadsheets, and embedded analytics. Cube has garnered a #1 ranking in Dresner Advisory Services' inaugural 2025 Semantic Layer Market Study for its data integration and access control strengths, modeling and transformation, and performance and optimization. A universal semantic layer is recognized as the "next big thing" in data architecture, providing an application-independent view of data objects that are critical to business operations. Cube Cloud combines the advantages of cloud-native technology with OLAP's structured methodology, filling the gap between flexibility and speed with a solid basis of trustworthy, controlled, and accessible data. The accuracy of underlying data is crucial for AI success, and Cube Cloud offers a scalable solution for companies searching to consolidate their data, improve decision-making, deploy successful AI initiatives, and eliminate friction from disparate platforms.

Fine-grained access control is essential for delivering the correct data to the right people. Data teams must ensure that users only see what they're supposed to see without slowing down development cycles or duplicating logic across dashboards, APIs, and applications. Cube Cloud's data access controls deliver a unified, governed, and scalable way to enforce data security across every data experience, including AI, BI, spreadsheets, and embedded analytics. This is achieved through query rewrite, Data Access Policies, LDAP integration, and Cube Cloud authentication, which allow organizations to centrally manage access at the row, column, and member level using flexible configuration in Cube's data model. Implementing data access controls helps keep sensitive information safe and ensures it's only used by the right people for the right reasons. Cube offers robust solutions for implementing granular security, including row-level, column-level, and member-level access control. These solutions ensure that sensitive information is only accessible to authorized users, enhancing data security and compliance within applications powered by Cube. Row-level security filters data based on user attributes, while column-level security restricts access to specific fields based on user roles or attributes. Member-level security controls the visibility of data model entities, such as cubes, views, and their members, using flexible configuration in Cube's data model. Data access policies can be used to configure these levels of security, allowing organizations to set up permissions so everyone gets the data they need for their job, but nothing more. By implementing these techniques, organizations can ensure secure, maintainable, and enterprise-ready access control strategies that protect sensitive information and meet compliance requirements.

This release of Cube Core version 1.3 includes various improvements to performance, data source support, and API endpoints. The update introduces breaking changes, including non-strict date range matches in pre-aggregations, removal of top-level `includes` and `excludes` parameters in views, a new API scope for the `/v1/sql` endpoint, updates to data source concurrency, and fixes to the `extends` parameter. Performance optimizations have also been implemented, such as deserializing Cube Store result sets in native Rust code and improving compilation speed with worker threads. Additionally, support has been added for rebuilding specific pre-aggregation partitions, SQL API queries in the `/v1/sql` endpoint, and a new subpath option for MinIO storage. The release also upgrades Node.js to v22 and introduces new features in the documentation.

The modern data stack is fragmented, leading to inconsistent reports, redundant data modeling, and debates over whose numbers are correct. This fragmentation occurs due to different teams defining key metrics differently, BI tools being siloed, repeated data modeling, varying query performance, and patchy data governance. A universal semantic layer can address these issues by standardizing metrics across every tool, centralizing data access controls, fastening scalable performance, and making the stack future-ready. With a universal semantic layer, executives trust dashboards, data teams reduce maintenance overhead, analysts explore data with confidence, product teams build smarter applications, and the entire organization operates from a single source of truth. Cube Cloud is purpose-built to be this universal semantic layer, connecting natively to cloud data warehouses, exposing consistent data models via APIs, and integrating seamlessly with AI, BI, spreadsheets, and embedded analytics.

We're thrilled to announce a significant leap forward in our AI API capabilities, particularly for our valued users leveraging Snowflake, with the introduction of the "Bring Your Own Large Language Model" (BYOLLM) feature. This groundbreaking addition empowers organizations to seamlessly integrate their own compatible Large Language Models (LLMs) with our AI API, directly addressing critical concerns surrounding data security and cost optimization in today's data-driven landscape. With BYOLLM, users can choose to use a compatible LLM from their cloud provider or now from Snowflake the AI Data Cloud, ensuring that sensitive data remains securely within the established Snowflake environment while unlocking the full potential of advanced AI capabilities. The integration offers significant cost management benefits by consolidating LLM usage within the existing Snowflake infrastructure and credit spend, eliminating the need for separate, potentially more expensive, AI services. Getting started with this game-changing integration is remarkably straightforward, involving enabling the BYOLLLM feature on the AI API configuration page within Cube Cloud and securely saving Snowflake Cortex credentials. We value our vibrant community of organizations using Cube or our fully managed Cube Cloud platform in conjunction with Snowflake and are committed to providing tailored support and guidance to facilitate a smooth and successful integration. The combination of Cube's AI API, BYOLLM, and Snowflake Cortex offers a powerful solution for organizations seeking to enhance data security, optimize costs, and unlock the transformative potential of AI.

Over the last decade, enterprises have struggled to scale data literacy programs that empower employees to understand data and make informed decisions. The rise of AI has created a new urgency, higher stakes, and greater hype, but organizations are again facing challenges in deploying generative AI across every function without good data. To address this, it's essential to build AI literacy by ensuring that every AI-generated answer is aligned with how the business defines success, using a universal semantic layer to govern data and provide context for AI applications.