March 2025 Summaries

Migrating from a legacy OLAP platform like Microsoft SQL Server Analysis Services (SSAS) can seem daunting, but with the right approach and Cube Cloud's universal semantic layer, businesses can migrate away from SSAS without disruption while unlocking cost savings, scalability, and improved performance. The key reasons for migrating from SSAS include cost savings, scalability, unified data access, and a future-ready stack. To migrate successfully, businesses should assess their current OLAP environment, plan the transition with considerations such as defining phases, engaging business users early, and setting up a test environment. Next, they should map their SSAS models to Cube Cloud's universal semantic layer, validate and optimize performance, and go live while monitoring adoption. By migrating from SSAS to Cube Cloud, enterprises can realize cost savings, faster query performance, seamless Excel connectivity, and future-ready analytics, supporting AI and embedded applications.

Microsoft SQL Server Analysis Services (SSAS), once a pivotal tool for enterprise Online Analytical Processing (OLAP), is increasingly seen as inadequate in cloud-first environments where platforms like Snowflake, Databricks, Amazon Redshift, and Google BigQuery dominate. SSAS struggles with elasticity, performance, and cost-efficiency, as its on-premises design leads to bottlenecks and high maintenance costs when integrated with cloud solutions. The need for seamless data accessibility across business intelligence tools is hindered by SSAS's isolated architecture, prompting organizations to seek alternatives. Cube Cloud emerges as a modern solution, offering a cloud-native semantic layer that supports Excel and Power BI, reduces infrastructure expenses, ensures compatibility with diverse data platforms, and enhances performance through intelligent caching. Transitioning to Cube Cloud allows businesses to modernize their OLAP strategies, ensuring agility and streamlined data access in cloud environments.

The struggles of Business Intelligence (BI) serve as a warning for the challenges that Artificial Intelligence (AI) may face when scaling. Despite significant investments and technological advancements, BI adoption has stalled due to issues such as data access not being enough, complexity overwhelming self-service initiatives, trust being fragile, people and processes lagging behind technology, and speed without strategy leading to waste. To avoid repeating these mistakes with AI, organizations need to focus on building a strong data foundation that ensures consistent, governed, and accessible data. A universal semantic layer is crucial in unifying data from multiple sources, standardizing business logic, and providing governed access to both AI and BI. By getting the data right, organizations can improve the success of their AI initiatives and avoid repeating the struggles of scaling BI.

The Gartner Data and Analytics Conference in Orlando highlighted the critical need for enterprises to balance speed with robust governance in the fast-evolving AI landscape, advocating for a modular, open, and collaborative approach to data management. The conference underscored the importance of trustworthy data and robust governance frameworks, emphasizing that investing time in solid governance can accelerate long-term innovation. The event also championed a best-of-breed approach in AI platforms, advocating for modular architectures that support integration and reusability, which are essential for next-generation data platforms. Despite technological advances, cultural roadblocks remain a significant challenge, with a strong data culture seen as crucial for maximizing AI investments. Cube's universal semantic layer, designed to unify and democratize data, aligns with these insights by providing a single source of truth and enabling collaboration across IT, data analysts, and AI agents, fostering an environment where data is a collaborative asset and driving innovation and operational excellence.

Data modeling is a crucial component of building the semantic layer, involving not only developers and data engineers but also data analysts and business users. Cube has embraced a code-first approach to developing data models but acknowledges the necessity for a no-code, visual method to accommodate non-technical users. As a response, Cube has launched Visual Model, a no-code data modeling experience in Cube Cloud that is now generally available, allowing non-technical users to engage in data modeling without coding while complementing the code-first approach to save time for experienced users. After months of refining its user interface and adding support for features like hierarchies and folders, Visual Model is now accessible to all Cube Cloud customers on Premium and above tiers, aiming to enhance team collaboration in data modeling.

The Cube AI API has significantly reduced its pricing for token usage due to the availability of more efficient models like Claude 3.5 Sonnet and GPT-4o, now allowing five requests per CCU for all Cube Cloud customers. The API, already enabled for paying customers, includes features like Value Search to improve data access and interaction by indexing specific dimensions for more effective queries. Customers can also bring their own LLMs to further customize and potentially reduce costs. Additionally, Cube has launched a Slack app to facilitate stakeholder interaction with the AI API, providing a seamless way for users to query and receive data through direct messaging, with options for follow-up questions and feedback.