How Perch Insights Powers AI-Driven CX Analytics with Cube Cloud

Perch Insights is a customer experience (CX) business intelligence platform that enables CX leaders to transition from reactive to proactive decision-making. By unifying data from multiple sources into a single source of truth, Perch Insights delivers actionable insights that improve customer lifetime value, reduce acquisition and service costs, and optimize the customer journey.

To enhance their AI-driven analytics capabilities, Perch Insights sought a universal semantic layer that could provide a consistent foundation for their dashboards, AI copilots, and CX-specific workflows. Cube Cloud’s universal semantic layer proved to be the ideal solution, delivering the flexibility, governance, and performance they needed.

The Challenges of Scaling Beyond a Single Interface

Before implementing Cube Cloud, Perch Insights faced several challenges in their data infrastructure. One of the primary issues was the lack of a headless semantic layer. While their built-in BI tools provided embedded dashboards, they did not support all of their requirements. This included being able to analyze the data for other applications beyond dashboards, specifically, their AI Copilot and AI-driven insights engine.

Additionally, Perch Insights needed a way to unify metrics across both their traditional dashboards and their AI-driven copilot. Without a single source of truth, they struggled to ensure consistency and reliability across these different analytical experiences.

Another major challenge was the need for client-specific data isolation. Perch Insights maintained separate data repositories for each client, ensuring that access remains strictly controlled and that data is never intermingled. However, this structure also introduced complexities in analytics workflows, requiring a semantic layer that could support multi-tenant analytics while enforcing data governance.

At the same time, Perch Insights wanted to leverage large language models (LLMs) to assist in querying and analyzing CX data, but they had concerns about allowing an LLM to directly define metric formulas. They recognized that raw SQL schemas lacked the necessary semantic context for AI-driven analytics and required a more structured approach.

Beyond these technical challenges, Perch Insights also struggled with the complexity of managing multiple semantic layers across their platform. Without a unified approach, they often faced duplication, inconsistencies, and extensive change management efforts to maintain accuracy. To build a scalable and AI-powered CX analytics platform, they needed a universal semantic layer that could address all these concerns while improving governance and performance.

Choosing Cube Cloud

As the team evaluated their options, they considered building an in-house semantic layer, integrating with the semantic capabilities of existing BI tools, and adopting dbt Lab’s evolving approach to metric management. Ultimately, they chose Cube because it provided the flexibility, governance, and performance they required. Cube’s open-source foundation, combined with the option to use Cube Cloud for scalability, made it an attractive choice that aligned with Perch Insights’ preference for extensibility and control.

In addition to deployment flexibility, Cube offered a well-structured metadata model that enabled Perch Insights to maintain strict governance over how metrics were defined and consumed. Cube offered flexibility to extend the semantic layer data model and add critical information about metrics, dimensions and their relationships needed for AI-driven analysis. This was critical for ensuring consistency across their AI-driven analytics and BI workflows. Performance optimization was another key factor in their decision.

Integrating Cube Cloud

The implementation of Cube’s universal semantic layer allowed Perch Insights to seamlessly integrate their analytics workflows. By leveraging Cube’s REST API for their embedded analytics application and SQL API for internal analytics, they connected Cube with different parts of their technology stack, enabling consistent metric definitions across all their analytics applications.

Additionally, Perch Insights developed custom data pipelines to validate and enforce alignment between their data warehouse, Cube Cloud, and CX applications. This ensured that every layer of their data infrastructure remained synchronized, reducing inconsistencies and maintaining a high standard of data integrity.

Cube also provided the flexibility Perch Insights needed to serve multiple use cases. Their AI copilot, dashboarding tools, monitoring systems, and modeling applications all relied on Cube Cloud as a common, trusted data access point. By standardizing query syntax and metric definitions, Cube made it possible for Perch Insights to provide a seamless analytics experience across various interfaces, whether for human or AI-powered decision-making.

“We knew we wanted to make use of LLMs to assist in querying specific data, but we had reservations about allowing the LLM to directly define a metric formula. Additionally, we felt like the semantics of a raw SQL schema were limited and would need supporting metadata.”

— Ryan Ward, VP of Engineering, Perch Insights

Transforming the Way CX Analytics Are Delivered

With Cube’s universal semantic layer in place, Perch Insights transformed the way they manage and deliver CX analytics. By consolidating their metrics into a governed, universal semantic layer, they eliminated inconsistencies and ensured that insights remained consistent across dashboards, AI copilots, and monitoring tools. This enabled them to scale their analytics capabilities without sacrificing governance or accuracy.

The adoption of Cube also allowed Perch Insights to maintain strict data security while supporting multi-tenant analytics. Each client’s data remained securely isolated, but at the same time, Perch Insights could apply standardized analytics models across all customers. This balance of security and scalability was crucial for delivering a reliable and efficient CX intelligence platform.

AI-powered decision-making also became significantly more effective with Cube Cloud. The universal semantic layer provided the metadata necessary for LLMs to query CX data intelligently, without introducing the risks of uncontrolled metric definitions. Additionally, Cube’s caching and pre-aggregation capabilities improved performance, ensuring that analytics queries ran efficiently and delivered insights in real time.

Leading the Way in CX Analytics

With Cube’s universal semantic layer, Perch Insights has successfully built a scalable and AI-driven CX analytics platform. By unifying dashboards, AI copilots, and monitoring tools under Cube Cloud’s universal semantic layer, they have created a consistent, high-performance analytics environment that ensures data integrity, governance, and flexibility. As AI continues to shape the future of CX intelligence, Perch Insights is well-positioned to lead the way, empowering CX leaders with trusted, scalable, and AI-optimized analytics solutions.