June 2025 Summaries

The emergence of agentic analytics, exemplified by Cube D3, is transforming traditional business intelligence by integrating AI-powered teammates with human efforts to enhance data exploration, model building, and insight generation. Departing from static dashboards and spreadsheet workflows, this approach leverages Cube Cloud’s universal semantic layer to ensure consistent and governed data outputs, allowing for efficient, natural language querying and rapid insight generation. The transition to agentic analytics involves identifying repetitive, high-impact use cases, establishing a strong semantic foundation, onboarding AI teammates, and continuously refining processes through user feedback and governance. By embedding AI agents into existing workflows, organizations can scale these benefits across teams and tools, fostering a culture of speed, collaboration, and autonomous data-driven decision-making.

The modern analytics stack, once considered the future, is now showing its age due to the limitations of traditional tools. The rise of generative and agentic AI is exposing these limitations, with business users still waiting for answers and data teams managing backlogs. Agentic analytics aims to bridge this gap by providing autonomous decision-making capabilities through the use of AI agents that generate queries, explore patterns, surface anomalies, and recommend actions grounded in context and driven by logic. Agentic analytics differs from traditional self-service BI tools, which provide more control but not always more insight, by leveraging generative AI techniques for data analysis. It bridges the gap between human roles in the data and analytics lifecycle, ensuring data integrity and automating exploration, decision-making, and follow-up tasks. With Cube D3's agentic analytics, users can ask questions into an Analytics Chat Interface, receiving governed semantic SQL queries and explanations within seconds, and receiving follow-ups and suggested next steps.

The future of analytics is shifting away from traditional dashboards, which have plateaued in adoption due to clutter, overwhelm, and lack of actionable insights. Instead, a new generation of analytics tools is emerging that focus on decision-centric workflows, powered by AI agents that can detect deviations, investigate causes, suggest actions, and automate follow-ups. These systems require a strong semantic foundation, such as Cube Cloud's universal semantic layer, to define business logic, ensure consistency, and govern data access. The goal is to propel users toward what to do next, rather than just presenting data, and to enable faster and more confident decision-making.

AI has entered the enterprise but not as a single all-knowing assistant managing everything. Instead, companies are realizing that generalist AI falls short in specialist environments and a new trend is emerging: domain-specific AI agents and assistants that deeply understand the workflows, terminology, data structures, and logic of a particular business function. These digital teammates know what your enterprise data teams need. Generalist AI tools often misinterpret requests, pull from the wrong data, or generate outputs that are directionally correct but operationally useless because context matters. In high-stakes business environments, close enough isn't good enough. Domain-specific AI agents understand common tasks, such as running time comparisons and generating dashboards, without needing retraining for every task. They know patterns, shared metrics, and filters, and can explain why numbers matter. Examples of this shift are seen in data engineers using AI to validate pipelines and finance teams using AI to reconcile numbers. Building domain-specific AI isn't just about training a model differently; it requires a different architecture. This is achieved with Cube D3, the trustworthy agentic analytics platform, which includes a semantic foundation, governance-first access, explainability and auditability, and interoperability. With these building blocks in place, Cube D3 operates alongside your team, within your rules, and in service of your business goals. Enterprise leaders are growing wary of one-size-fits-all AI and need domain-specific solutions that offer better answers to questions like trustworthiness, integration, ownership, and scalability. Domain-specific AI agents extend people by automating what's tedious, accelerating what's manual, and enhancing what's strategic. The most successful AI deployments in the enterprise won't be general copilots but rather digital teammates that act like domain-specific AI agents.

Cube D3, our first-party agentic analytics platform, is here and represents an evolution of Cube, with a new analytical interface for users to interact with. It's powered by the Cube semantic layer and builds upon the existing core pieces of the Cube Cloud platform, including semantic data modeling, access control, caching, and API endpoints. The introduction of D3 doesn't change Cube's open platform approach, which enables customers to arrange the pieces to best support their goals. It will initially support inference from major partners such as Snowflake, Databricks, Amazon, Microsoft, and Google, with plans to add more inference partners as requested by customers. The platform will gradually roll out access throughout the summer, with general availability planned for the fall. Joining the D3 waitlist is encouraged.

D3, an agentic analytics platform from Cube Cloud, tackles the challenge of providing unified access to insights by leveraging a robust semantic layer. This layer serves as a bridge between complex data structures and business insights, empowering both data consumers and stewards with augmented self-serve capabilities. By embedding metadata and standardizing business logic, D3's semantic layer provides a single source of truth for business metrics and dimensions, ensuring consistent definitions and reducing ambiguity. The platform introduces analytics agents that assist users in interacting with data using natural language, eliminating technical bottlenecks and driving insight velocity. With its key components, including agent spaces, rules, and certified queries, D3 is designed to provide a transformative opportunity for organizations to make sense of their increasing data complexity, governance hurdles, and the difficulty of providing unified access to insights.

Today, the company is launching D3, an agentic analytics platform powered by the Cube semantic layer. The launch comes after building Cube over 5 years ago as a side project to power a Slack AI chatbot for analytics. Thousands of organizations now use both open-source and cloud products to power business intelligence and embedded analytics workloads. D3 is unique because it was built from first principles for AI-augmented workflow and is fully based on semantic understanding of data. It introduces several core experiences, including Analytics Chat, Workbooks, Data Apps, and Semantic Modeling, all built around the Cube semantic layer and Semantic SQL. The platform aims to provide a structured, governed way to query and extend the semantic layer, ensuring semantic layer definitions are used correctly. D3 is designed for both humans and AI agents, with a multi-agent ecosystem envisioned where sales, marketing, logistics, and other AI agents can seamlessly request insights from D3 agents. The company plans to roll out access to users throughout the summer with general availability in the fall.