Lepaya Launched Enterprise-Ready Embedded Analytics in Three Months with Holistics

In June 2024, Rob Winters, leading data at Lepaya, had a problem: he had eight weeks to find an embedded analytics solution or renew a license for a legacy BI tool that, while strong on visualizations, no longer aligned with the direction he wanted to take.

Lepaya, a fast-growing L&D platform serving clients like Dell and KPMG, needed to give enterprise customers access to their own learning data. However, the limitations of their incumbent visualization platform — rigid embedding, licensing constraints, and lack of customization — made it a poor fit from the start.

That’s when he turned to Holistics, a tool he had worked with years earlier while supporting over 1,000 users at a prior company. Back then, he had used Holistics for self-service analytics, not embedding, but he knew its strengths: flexible modeling, fast development, and responsive support. With embedding as the priority, he decided it was time to give Holistics another look. And Holistics did not disappoint.

Three months later, Lepaya had a fully functional Embedded Dashboard Portal that felt native, scaled cleanly, and had virtually zero dependency on frontend engineers for dashboard changes.

When Lepaya set out to build a customer-facing analytics portal, Rob Winters knew exactly what he needed. The solution had to meet a principled checklist of requirements:

• A clean, native-feeling embedded experience
• Row-level security and flexible filtering
• Developer control over the embedded behavior
• Built-in self-service features like drilldowns and exports
• Strong governance and usage visibility
• A consumption-based pricing model that could scale gradually

Lepaya’s incumbent BI tool, a leading visualization platform, was the obvious starting point for a quick proof of concept. But it didn’t take long to see it wasn’t the right fit.

The platform was built for traditional dashboard delivery, not deeply integrated product experiences. While it served internal reporting needs well in the past, the team found it difficult to tailor the embedded experience to meet their design and interaction standards.

The cost model also posed another challenge: the tool’s seat-based pricing meant they had to pay for each potential user, regardless of the actual usage frequency.

That led to a deeper question: should they just build it themselves?

On paper, building the embedded module in-house looked tempting. They had a strong team and a clear idea of what they wanted. But Rob quickly recognized the trade-offs. “We have a fairly small engineering team, so investing time in something that isn’t on the critical path is tough. Building obviously slows iteration. You start building technical dependency."

Building it right would take months. They needed a polished, scalable embedded experience that would match the rest of Lepaya’s product. So Rob and his team went on to benchmark five vendors against his criteria. Holistics came out ahead in nearly every category.

Holistics changed the cost-benefit conversation. With Holistics, Rob and his team got what they wanted without burning months of engineering time or putting up with unscalable pricing models that other third-party BI tools provide.

Holistics allows product engineers to integrate analytics into their products without compromising on development best practices

Once Lepaya chose Holistics, implementation moved fast. The implementation team was small, just Julian, a product data analyst, and one developer, but they delivered a full embedded analytics portal for 15 enterprise clients within a few months.

Since launching their embedded analytics portal with Holistics, Lepaya has seen a steady increase in client engagement.

More and more of their enterprise customers are logging in regularly to extract insights about how their teams are engaging with training programs, and whether those programs are working.

For Lepaya, it was critical that the embedded analytics portal feel like a native part of the product, and not an awkwardly layered-on afterthought. Julian and the team knew how difficult it could be. They had worked with tools that made this nearly impossible: rigid embedding logic, limited design control, and developer-hostile workflows. Holistics was the opposite.

A big reason was Canvas Dashboards, which reimagine dashboards as collections of code-based "Blocks" (e.g, Visualization blocks, Control blocks, Text blocks) that can be freely arranged on a blank canvas. Analysts can apply custom styling through code to ensure every block fits seamlessly with the product’s design system. Each block can also be parameterized with variables and reused across dashboards, making it easy to maintain consistency at scale.

With Canvas Dashboards and Custom CSS, Lepaya’s product design team was finally free to shape the analytics experience like any other part of Lepaya’s app.

During implementation, one of the things that stood out to Julian was how Holistics’ implementation of analytics-as-code strikes a rare balance between UI convenience and code-level control, offering him the freedom to move quickly, whether he was building new datasets, duplicating complex logic, or iterating on metrics across dashboards.

As an experienced SQL practitioner, Julian quickly saw the value of Holistics AQL (Analytics Query Language) as it allows him to take a metric-first, composable approach to analytics.

AQL allows analysts to define a base metric once, then reference, filter, and transform it in downstream logic.

With AQL, Julian can define a base metric once, then reference, filter, and transform it in downstream logic, reducing duplication and making models and dashboards easier to maintain. Two specific features enable this style of work: the pipe operator, and dimensionalize().

The pipe operator allows Julian to structure multiple expressions into a clear, sequential pipeline, where the output of one expression becomes the input for the next. This makes it easy to layer filters, transformations, and calculations without nesting or repeating logic. It’s faster to write, easier to debug, and far more maintainable.

The dimensionalize() function complements this by letting Julian reuse a metric across different groupings without redefining the logic each time. Rather than rebuilding the same calculation for every reporting view, he can apply a consistent metric and just switch the dimension on top. Together, they turn metric modeling into a modular, composable process.

After seeing how fast they could ship a polished embedded experience for clients, the team asked: What if we used the same platform to reduce internal reporting overhead?

There had been talk of self-service before. The engineering team had experimented with Metabase, but adoption stalled outside of technical users. It wasn’t built for people who don’t write SQL, and company-wide self-service doesn’t work if only a few can use the tool.

Adding to the challenge was Lepaya’s B2B learning model, which brought a high degree of variability.

“Every customer is getting a slightly different product experience and delivery experience,” the team explained. “So you get a lot of very unique questions you have to solve, and you have to build multiple separate dashboards. That makes your life absolutely living hell from a BI standpoint.” Standardized dashboards just couldn’t keep up. “We tried that with our previous BI tool. We’d end up with literally 30+ filters on a single dashboard,” Rob said. “And at the end, the user still exports the data to Google Sheets to make their own chart.”

After the success of embedded analytics, it became clear that the same platform could also solve their long-standing self-service challenges.

Rob and his team saw Holistics as an opportunity to drive a different kind of data culture where, instead of building hundreds of reports to chase every edge case, they could take a different approach: expose curated datasets, and let the people closest to the problem explore the data themselves.

The team didn’t assume that rolling out a self-service BI tool would be enough on its own. They knew that just making data available doesn’t guarantee adoption, so they invested early in enablement.

We did kickoff calls and live training sessions,” Julian said. “We always began by asking: ‘What are some things you want to learn?’ And then showed them how to do it. That first aha moment is what matters.” They also created detailed documentation for each dataset—living Confluence pages with walkthroughs, videos, and clear explanations. Combined with Holistics’ intuitive interface, these resources helped non-technical users get value immediately. “It was super helpful,” they added. “We could finally get people started on actual use cases that mattered to them.”

That effort paid off, especially for the Customer Success team.

In the past, CSMs relied on analysts to answer client-specific questions: who attended training, how did they rate it, or what’s the latest engagement trend? These were often one-off requests that required drilling into specific datasets. With Holistics, the team now handles many of these questions on their own.

Now, more teams have begun exploring datasets, answering their questions on their own. Customer Success no longer depends on Slack threads for every client request. Product managers can drill into feature usage without waiting for custom reports. And the Finance team benefits from a familiar, tabular data experience they can easily analyze and export.

Today, Holistics continues to evolve alongside Lepaya. Rob and the team acknowledged that some parts of the experience still feel rough, but they’ve also seen meaningful progress, shaped by their own feedback. That responsiveness has laid the foundation for a strong, collaborative partnership going forward.

“I’ve said this a thousand times, but I’ll keep saying it,” Rob said. “This is so different from working with other BI companies. Here, it’s clearly all about the people. It comes through in the product, and it comes through in the engagement.”