ChatGPT
Perplexity
.png)
.png)
Build your first embedded data product now. Talk to our product experts for a guided demo or get your hands dirty with a free 10-day trial.
.png)
Recently, the team at ClickHouse published a story about how Luzmo uses ClickHouse Cloud to power fast end-user analytics. It reflects a journey that started nearly eight years ago and has become an important part of the infrastructure behind our platform.
While the article focuses on the ClickHouse side of the story, we thought it would be useful to share a bit more context from the Luzmo perspective: what problems we were trying to solve, why ClickHouse stood out early on, and how our architecture evolved as embedded analytics workloads grew.
If you’re building a data-heavy product, many of these challenges will likely sound familiar.
At Luzmo, our goal is simple: help software companies embed powerful analytics directly into their products.
But delivering analytics inside an application is very different from running dashboards for internal teams.
Embedded analytics platforms need to handle:
In many cases, our customers don’t know exactly what questions their users will ask until the data is already flowing through the system. That means the underlying data infrastructure must be extremely flexible and performant at the same time.
Early on, we realized traditional approaches to analytics infrastructure would struggle with this kind of workload.
When we first encountered ClickHouse around 2017, it immediately stood out.
At the time, achieving high-performance analytics often meant operating complex distributed systems or large clusters. Performance improvements typically came with significant infrastructure overhead.
ClickHouse offered a different balance.
It delivered impressive query performance even at a smaller scale while still offering a clear path to scale as data volumes grew. For a company building a platform designed to process large volumes of analytics data for many different customers, that combination was compelling.
As our co-founder and CTO Haroen Vermylen said: “ClickHouse worked on my laptop. That was already pretty amazing.”
From that point forward, ClickHouse became an important component of how we powered analytics workloads internally and for our customers.
For several years we… operated ClickHouse ourselves.
Running infrastructure gave us flexibility, but it also came with operational overhead. Managing clusters, handling upgrades, and maintaining performance all required time and attention that we would rather spend improving the analytics experience for our users.
When ClickHouse Cloud became available, the value proposition was clear.
Moving to a managed service allowed our engineering team to offload infrastructure management while keeping the performance and scalability benefits we relied on. Instead of running clusters, we could focus on building better analytics capabilities inside Luzmo.
That shift allowed us to move faster and dedicate more engineering effort to what matters most: the product experience.
What started as a database powering analytics queries gradually expanded to support additional workloads inside Luzmo.
Today, ClickHouse is used in several ways across our platform.
For example, we store query logs and plugin logs to better understand how analytics queries behave in production. This visibility helps both our team and our customers diagnose performance issues and optimize data models.
More recently, ClickHouse has also become part of our work with vector search and AI-driven analytics experiences. Rather than introducing separate infrastructure, we’ve been able to extend our existing platform capabilities.
As Haroen jokingly puts it: “ClickHouse has been quite infectious inside our company.”
Once it’s part of the architecture, it often becomes the most practical solution for the next data challenge.
One of the unique aspects of embedded analytics is that you often have to engineer systems that handle unknown future workloads.
Because Luzmo serves many different software companies, we rarely know in advance what data structures will look like. Our platform regularly processes:
To manage this complexity, we built a query engine that can optimize and coordinate requests across multiple sources. In many cases, large numbers of incoming queries can be consolidated into a smaller set of optimized operations.
This architecture allows us to deliver fast analytics experiences while maintaining efficiency in the underlying data infrastructure.
Ultimately, the goal of embedded analytics is simple: analytics should feel like a natural part of the product.
Users shouldn’t think about databases, query engines, or infrastructure. They should simply explore their data and get answers quickly.
And Luzmo allows that. By integrating powerful visualizations and AI-driven exploration directly into the application, Luzmo helps teams building data-centric software turn high-volume information into intuitive, actionable insights for their customers.
Delivering that experience requires infrastructure that can handle large volumes of queries while remaining flexible enough to support constantly changing data models.
That’s why performance, versatility, and cost efficiency continue to be critical characteristics of the technologies behind the platform.
Our collaboration with ClickHouse has been an important part of the journey toward making embedded analytics faster and more scalable for modern SaaS products.
If you're interested in the ClickHouse perspective on this collaboration, you can read their full article here:
How Luzmo uses ClickHouse Cloud to power fast end-user analytics https://clickhouse.com/blog/luzmo
And if you're building a product where analytics needs to be delivered directly to your users, we’d love to show you how Luzmo can help.
All your questions answered.
Build your first embedded data product now. Talk to our product experts for a guided demo or get your hands dirty with a free 10-day trial.
