Name: Answers without the ticket
Brand: Cube

Brex


Brex is an intelligent finance platform that combines corporate cards, spend management, expense tracking, business banking, and travel into a single product, serving over 35,000 companies globally. As Brex set out to redefine financial reporting for its customers, they needed a foundation that could power AI-native experiences with the accuracy, governance, and scale that finance teams demand.

<video controls width="100%" preload="metadata">
  <source src="https://staticassets.goldcast.io/organization/ec59e10f-fc8c-4594-950b-a2cd4d3aba9c/videos/8f54e8d0-c6f9-4adf-84d6-1daf5f6b9568_prod.mp4" type="video/mp4" />
</video>

### The Challenge: Moving Beyond Static Dashboards

Brex saw that financial reporting across the industry hadn't fundamentally changed in two decades. When a CFO asks "Why is marketing 23% over budget this quarter?", the dashboard shows the *what* but never the *why*. Getting a real answer typically requires filing a ticket, waiting for an analyst to write SQL, validating the output, and pasting results into a slide — a two-second question turning into a two-day project. Brex set out to eliminate that gap for every finance team using their platform.

This created several compounding problems Brex needed to solve:

- **Rigid dashboards, inflexible answers:** Pre-built dashboards are fast but only answer pre-defined questions. The moment a finance leader asks something new, the entire ad-hoc cycle restarts.
- **Bottlenecked data teams:** Analysts spend significant time on repetitive reporting requests instead of strategic work, while business users wait days for answers they need in real time.
- **No interpretation layer:** Charts present data but don't explain anomalies, surface historical context, or recommend what to look at next — the things a senior analyst would naturally do.
- **In-house AI prototype hit a wall:** Brex initially built a lightweight in-house agent on top of self-hosted open-source Cube Core. Maintaining the agent required heavy infrastructure overhead — blue-green deployments, on-call rotations, and reliability work. Their prompt layer ballooned to over 3,000 lines across four files, creating a maintenance treadmill where every new capability risked breaking another.

Brex needed a way to give every customer the equivalent of an embedded AI financial analyst — one that understood Brex-specific concepts like card programs, expense policies, entity structures, and approval workflows — without rebuilding the underlying analytics infrastructure from scratch.

### The Solution: An AI-Native Reporting Stack Powered by Cube

Brex built **Spaces**, an AI-powered workspace for finance teams that delivers insights in seconds rather than days. Underpinning Spaces is a deliberate architectural decision: invest in ontology and semantics before writing a single line of UI code.

After evaluating semantic layer providers including Cube, dbt Semantic Layer, and LookML against AI-readiness criteria, Brex chose Cube and ultimately migrated from their self-hosted open-source deployment to Cube Cloud. The decision came down to three factors: eliminating infrastructure overhead, replacing thousands of lines of brittle prompts with a managed agent, and — most importantly — evaluation results showing that Cube's cloud agent significantly outperformed their in-house agent on data consistency, tool usage, and answer quality.

Cube became the foundation of a three-layer context engineering approach:

1. **The proprietary semantic layer** encodes Brex's financial domain — what "spend" means, how departments roll up, how entity structures map to accounting hierarchies, how compliance thresholds are defined. This is what turns a generic SQL agent into a Brex financial analyst, and what makes Spaces a Brex product rather than a generic BI tool.
2. **Certified Queries** are pre-validated, analyst-approved SQL patterns the AI references for the most common financial questions. Cube's Certified Queries feature lets Brex curate gold-standard logic that the agent adapts from, rather than generating queries from scratch every time — dramatically reducing hallucinations and increasing trust.
3. **Agent Rules** shape behavior at the system level — when to flag anomalies, how to contextualize against historical patterns, how to format insights. Cube's Agent Rules are declarative instructions configured directly in Cube Cloud (with both *Always* rules that apply to every interaction and *Agent Requested* rules the agent invokes contextually). By moving to Cube Cloud, Brex replaced 3,000+ lines of custom prompts with roughly 10 Agent Rules unique to their business, letting Cube handle orchestration, querying, and chart generation.

To ship AI-native reporting to customers with confidence, Brex also built a robust evaluation system on top of Cube. **Online evaluators** run lightweight, deterministic checks for production edge cases — like distinguishing "my travel spend" from "the account's travel spend," or ensuring the agent discovers exact enum values before running SQL filters. **Offline evaluations** use LLMs as judges to score replays across answer quality, chart appropriateness, data consistency, insight relevance, and semantic layer usage. This continuous loop helped Brex raise their insight relevance score from the high 50s to nearly 90%.

<Testimonial />

### The Results: From Charts to Insights, From Days to Seconds

By building Spaces on Cube's semantic layer and AI-native infrastructure, Brex transformed financial reporting from a backward-looking artifact into a real-time analyst experience:

- **Two-day questions answered in two seconds:** Finance teams using Spaces no longer wait for analyst tickets to understand budget variances, spend anomalies, or compliance flags — they get explanations instantly, with charts, context, and recommended follow-ups.
- **Insights, not just data:** Spaces doesn't just visualize numbers — it surfaces anomalies, contextualizes results against history and benchmarks, and explains in natural language what happened, why it matters, and what to look at next.
- **Massive engineering simplification:** By moving from a self-managed open-source deployment with 3,000+ lines of prompt code to Cube Cloud, Brex eliminated infrastructure on-call burden and reduced their custom guidance to roughly 10 Agent Rules configured natively in Cube.
- **Higher answer quality, measurably:** Cube's managed agent outperformed Brex's in-house agent across data consistency, tool usage, and overall answer quality in head-to-head evaluations, while their evaluation system pushed insight relevance scores from the high 50s to nearly 90%.
- **A defensible, multi-tenant foundation:** Brex's investment in proprietary semantics, Certified Queries, and continuous evals creates a product moat that compounds over time — one that scales across thousands of customers, each with isolated data but a shared, governed model.
- **Conversational, refinement-friendly UX:** Internal user testing revealed customers don't want a single answer — they want to refine, filter by date, drill into segments, and compare periods. Cube's semantic layer supports this conversational pattern natively, making Spaces feel less like a dashboard and more like working with an analyst.

Brex's broader takeaway, applicable to any team building AI-native data products: ontology comes before UX, Certified Queries are how you earn trust, and evaluation systems are non-negotiable when customers make real financial decisions on your AI's output. As Brex puts it, the LLM is commodity — the context is the product.


Sign up Now

Brex built Spaces, an AI-powered finance workspace, on Cube's semantic layer. By replacing 3,000+ lines of in-house prompts with Cube Cloud's managed agent, Certified Queries, and Agent Rules, Brex turned two-day reporting requests into two-second insights for every customer.

How Brex built Spaces, an embedded AI financial analyst for 35,000+ customers, on Cube Cloud's universal semantic layer.

Replaced 3,000+ lines of in-house prompts with Cube Cloud — answers in seconds, not days

The future of reporting isn't a chart, it's an insight. Large language models are becoming a commodity — the LLM is the engine, but the semantic layer is the map. A well-modeled ontology is the difference between 'I don't understand that question' and a correct, contextualized answer with a chart and a clear explanation. Cube gives us the foundation to make that real for every customer.

Brex builds an embedded AI financial analyst for 35,000+ customers with Cube

Embedded Analytics

Build consistent, secure, and performant embedded analytics.

Drata


Drata provides an AI-native trust management platform, automating and centralizing governance, risk, compliance, and assurance for thousands of businesses globally, including a third of the cloud 100. As their customer base grew and their data platform matured, Drata needed to deliver real-time insights both internally to their teams and externally to their customers and partners.

### The Challenge: Scaling Analytics and Streamlining QBRs

Drata faced the critical challenge of scaling its analytics capabilities without scaling its team linearly (scaling without adding proportional headcount).

Drata's Customer Success Managers (CSMs) were burdened by time-consuming manual processes:

- __Manual QBR Generation:__ CSMs needed to create data-driven Quarterly Business Review (QBR) presentations for every major account, a process that typically took 2 to 3 hours per QBR. Multiplied across accounts, this resulted in CSMs spending 20 to 30 hours per quarter—almost a full work week—just building presentations.
- __Inconsistent Data:__ The manual process required CSMs to pull metrics from up to five different dashboards, calculate benchmarks by hand, and then format the data into slides and build a narrative. Disconnected data sources often resulted in inconsistent metrics, which eroded client trust.
- __Multi-Instance Management:__ Clients who manage 10 to 20 Drata client accounts (like Managed Security Service Providers) needed a unified dashboard. Previously, they had to log in and out of each account individually to check key metrics, which was exhausting and time-consuming.

### The Solution: AI-Driven Analytics
Drata adopted a dual approach to Cube, using one consistent semantic layer to power two distinct consumption patterns: structured API queries and AI-powered natural language integration. Cube became their single source of truth, providing a consistent, governed foundation for analytics that scales.
Crucially, the success of the AI-powered solution relied on Drata’s investment in an AI-focused semantic layer design. This involved ensuring clear, descriptive dimension and measure names, rich metadata, explicit relationships, and pre-calculating advanced metrics (like segment benchmarks).

> “__Cube becomes our single source of truth for metric definitions and powers everything from customer-facing dashboards to AI-driven quarterly business reviews.__ CSMs gain back dozens of hours each quarter and are equipped with everything they need to deliver intelligent storytelling, enabled by Cube’s semantic layer and agentic analytics. This represents a shift from static reporting to intelligent storytelling.”

— Anthony Cronander, Senior Analytics Engineer, Data

### The Results: Intelligent Storytelling and Time Savings
By leveraging [Cube](https://cube.dev/), Drata achieved significant operational improvements, transitioning from manual effort to automated, strategic insights:

- __Time Savings:__ CSMs get dozens of hours back in a quarter. The previous manual process consumed 20–30 hours per CSM per quarter
- __Shift in Output:__ The workflow creates a shift from static reporting to intelligent storytelling. Cube generates data-driven narratives and insights written in natural language, not just numbers
- __Strategy & Insight:__ Cube surfaces retention risks and expansion opportunities, comparing metrics to benchmarks and identifying outliers
- __Customer Trust:__ Context injection allows Cube AI to prioritize relevant metrics based on customer concerns. The resulting QBRs feel tailored, not templated, demonstrating deep knowledge of each account, which builds trust and guides the relationship to a strategic partnership
- __Scalability:__ Drata can build once and consume many ways (APIs, AI agents, embedded analytics) from a single rock-solid semantic foundation, ensuring consistency across all consumption layers


Drata scaled analytics using Cube’s AI-focused semantic layer as the single source of truth for metric definitions, saving CSMs dozens of hours quarterly and shifting customer success to intelligent storytelling.

CSMs gain back dozens of hours every quarter on QBR generation and customer-facing dashboards

Cube becomes our single source of truth for metric definitions and powers everything from customer-facing dashboards to AI-driven quarterly business reviews. CSMs gain back dozens of hours each quarter, enabled by Cube’s semantic layer and agentic analytics.

Drata scales data-driven decision-making with Cube's semantic layer and AI agents

Business Intelligence

Business Intelligence keeps your data in-sync

Webflow


Webflow is a Website Experience Platform (WXP) that empowers modern marketing teams to visually build, manage, and optimize stunning websites. Webflow’s site builder, CMS, and optimization tools put the power of code in a visual-first front end—where marketing and design can self-serve and engineering can extend with custom code. Webflow has over 300,000 customers in 200+ countries.

Webflow faced a growing need to streamline its data infrastructure. As the company scaled, it became essential to provide consistent data access and interpretation across both engineering and non-engineering teams inside and outside the organization, while maintaining a high level of performance and security. Rocio Delgado, Senior Engineering Manager‬, shares Webflow’s story.

## The Challenge of Providing Consistent, Scalable Access for Internal and External Use Cases
Before Cube Cloud, Webflow's engineering team contended with the challenge of providing scalable, centralized data access, not only for the product they were building but also for various internal analytics. [Webflow’s site usage dashboard](https://help.webflow.com/hc/en-us/articles/33961410031507-Bandwidth-overview) is a customer-facing Analytics dashboard that provides information about bandwidth usage and high-bandwidth pages and assets. 

The data that powers this dashboard, needs to be available internally too for billing and offline analysis. With multiple teams relying on data for both use cases, there was a pressing need for an abstraction layer that could offer uniform data interpretation.

Webflow sought a way to provide query scalability while simplifying data access. The primary goal was to develop a universal semantic layer that could define business logic consistently across the organization. This solution had to be scalable, secure, easy to implement, and cost-effective, while also supporting Webflow’s existing API infrastructure.

## Why Webflow Chose Cube Cloud
Webflow chose Cube Cloud for performance, flexibility, ease of use, and scalability. The decision to implement Cube Cloud was driven by its ability to power real-time data dashboards and its compatibility with Webflow’s tech stack. Cube’s [performance features](https://cube.dev/product/caching-and-data-performance), such as query caching, provided the speed necessary for responsive analytics. 

Webflow successfully integrated Cube Cloud into its infrastructure, pairing it with ClickHouse for high-performance querying. Webflow achieved a p95 of 220 milliseconds using Cube Cloud’s advanced caching capabilities with ClickHouse. Cube Cloud helped the team to ensure streamlined version control and easier collaboration between teams. 

![](https://ucarecdn.com/5915ea8d-1661-4d57-a064-97fef6b0d1ed/-/quality/smart/-/format/auto/)

Additionally, Cube's flexibility with multiple data sources and its ease of use—particularly in defining complex metrics and access through REST and GraphQL APIs—made it the perfect fit for Webflow’s needs. Cube’s [Data APIs](https://cube.dev/product/data-apis) allowed Webflow to easily integrate with various front-end‬ applications and third-party tools. This flexibility has made it easier to build and maintain data-driven applications.‬

> “We integrated Cube Cloud smoothly with ClickHouse, leveraging both for fast query‬ execution while maintaining the abstraction needed for different teams to access data‬ without diving into database-specific complexities.‬”

— Rocio Delgado, Senior Engineering Manager, Webflow

## Webflow Delivers Unified Data with Cube Cloud
Since implementing Cube Cloud, Webflow has seen a marked improvement in data accessibility and governance. Real-time analytics development timelines have been slashed, thanks to Cube Cloud’s architecture and native APIs, which eliminated the need for custom-built API layers. 

Enhanced speed and scalability have enabled Webflow’s teams to focus on innovation without being bogged down by data complexities. Cube Cloud’s universal semantic layer has not only improved Webflow's internal data processes but also empowered the organization to deliver more consistent and efficient data-driven experiences for its users. 

Cube Cloud’s universal semantic layer transformed how Webflow accesses and uses data. By centralizing business logic, it eliminated inconsistencies across teams and allowed for rapid deployment of new data models. The REST API integrations facilitated smooth data flows into Webflow’s front-end applications, simplifying the build and maintenance of data-driven features.


Ready to upgrade your data stack?

Customer-facing site usage analytics and internal billing analysis from one semantic layer

We integrated Cube Cloud smoothly with ClickHouse, leveraging both for fast query‬ execution while maintaining the abstraction needed for different teams to access data‬ without diving into database-specific complexities.‬

Webflow delivers internal and external embedded analytics with Cube Cloud

Alcon


Alcon is the global leader in eye care products. Alcon sought a solution that would provide a unified data modeling and access framework. Their vision was to use conventional dashboards augmented by Agentic AI to enable self-service analytics and drive actionable insights

## The Challenge: Discrepancy, Delay, and Digital Risk

**Alcon, like many large enterprises, faced challenges stemming from pervasive dashboards and complex data structures:**

- **Metric Inconsistency:** The same KPIs and metrics could have different definitions and calculation methodologies across various teams, leading to discrepancies and confusion.
- **Generative AI Risk:** Due to complex data schemas with many databases, tables, and fields, implementing GenAI posed a high risk of error and hallucination. Alcon also needed to maintain strict compliance with regulations like HIPAA and GDPR, requiring robust security
- **Slow Time-to-Value:** Users constantly requested questions that existing dashboards could not answer. Submitting new development requests, even using an Agile process, resulted in slow intake, development, and deployment cycles

## The Solution: Cube's Unified Semantic Layer and Agentic Analytics Platform

The selection of **[Cube Core](https://cube.dev/product/cube-core) as their semantic layer provider** was based on four core capabilities that aligned directly with Alcon’s requirements:

- **Access Control:** Cube enforces centrally managed data security and role-based access necessary for a highly regulated environment
- **Centralized Data Model:** Cube facilitates the centralized definition and management of business logic and KPIs
- **Caching:** Cube minimizes repetitive queries into the data lake/warehouses, performing aggregation to achieve faster performance
- **APIs:** Cube’s APIs allow seamless integration with downstream BI tools and, critically, with the Cube Agentic Analytics platform

![](https://lgo0ecceic.ucarecd.net/71abc7c1-8041-46e7-8d82-6ae77415a0f2/)

By implementing Cube, Alcon defined multiple logical data representations—**Cubes** (for accounts, products, orders, etc.) and **Views** (logical combinations that expose data fields)—to centralize metric management and calculation definitions in one place

> “__Semantic SQL is a wonderful thing – it's really at the heart of the semantic layer.__ 
Without Cube’s semantic SQL, our data analysts might have to write 20 different queries for a single core business metric. With Cube, that metric is defined once in the data model, and every downstream tool uses that definition along with the associated calculation logic.”

— Dr. Jun Huang, Global Head of Data Science for Digital Health, Alcon

## The Impact: Governance, Trust, and Accelerated Insights

Implementing Cube’s semantic layer and the [Cube platform](https://cube.dev/product/cube) delivered immediate benefits, enabling semantic governance in action.

- **Consistency and Trust:** The unified definitions provided by Cube ensure that data consumers receive consistent results. When testing the new system, the GenAI chatbot's responses to KPI questions matched exactly what was displayed on the existing Tableau dashboard. This consistency ultimately built user trust
- **Reduced Development Effort:** The Agentic AI complements the dashboard, allowing users to ask questions beyond the dashboard's scope. This capability reduces the need to develop as many dashboards as before
- **Improved Governance for AI:** The well-constructed semantic layer brings the correct context to the AI agents, significantly reducing the risk of hallucination. Furthermore, Alcon uses the Agentic Analytics platform to set rules, ensuring the agent only provides responses using approved metrics, adding a necessary level of control


Alcon leveraged Cube's semantic layer and Agentic AI to overcome metric inconsistency and AI risk in global operations. Cube enabled consistent, fast self-service analytics that answers questions beyond dashboard scope

Consistent self-service analytics across global operations, with GenAI grounded in the semantic layer

Without Cube, our data analysts might have to write 20 different queries for a single core business metric. With Cube, that metric is defined once in the data model, and every downstream tool uses that definition along with the associated calculation logic.

Alcon transitions from dashboards to dialogue with Cube's agentic analytics

Intuit

Tubi

Freshworks

Business users self-serve in natural language, analysts go faster, and every answer stays grounded in the governed semantic layer — for your team and your customers.

Self-serve analytics

Answers without the ticket

Describe the analysis and the agent drafts the query, chart, or report — while you keep control of every field, filter, and label.

Ask the Workbook Agent

Start in natural language, then drop into point-and-click, raw SQL, or Semantic SQL on governed metrics when you want more depth.

Three ways to query

Every chart is backed by a reusable, governed report, so an ad-hoc exploration stays consistent with the rest of your analytics.

Governed by construction

Explore in a workbook, with an AI analyst beside you

Break a metric down, extend the range, or ask what a definition means — a follow-up, not a new ticket, on the same governed model.

Ask the Dashboard Agent

Chart, KPI, filter, time-grain, text, and AI Summary widgets, arranged the way your team reads them.

Every widget you need

Scheduled refreshes, PNG and PDF snapshots, and embed-anywhere — so the answer reaches whoever needs it.

Self-serve, then share

Drill into a dashboard by asking

Analytics Chat, workbooks, and dashboards — explore, build, and publish without leaving the platform.

In Cube

Query the governed model from Claude, ChatGPT, or an agent you build, each connected over MCP.

In your AI tools

Drop the Slack Agent into a thread and follow up where your team already talks — permissions carry through.

In Slack

Explore where you already work

More than 400 companies ground their exploration on Cube's governed model, so answers stay consistent and scoped to each user — your team and your customers alike. Brex chose Cube over the dbt Semantic Layer and LookML.