Since we started Lightdash, users have wanted a way to make sure ad-hoc exploration work wasn’t just that - adhoc. They wanted an easy way to do the exploratory work they needed to in SQL, then a really simple UX to push that work back to dbt once they were done.
Today, this is easier than ever to do in Lightdash with the launch of our on-demand data modelling: a new way to develop data models with Lightdash and dbt that’s faster and more streamlined than ever before.
You can now:
We’ve built lots of new features that feed into this improved workflow of developing your data models in Lightdash. Keep reading to learn more about these new features and how we got here, or you can skip ahead and watch the demo.
Before Lightdash, we were working as data consultants helping teams build out their data stacks.
On every project, when it came to picking a BI tool, we ran into the same issue:
we loved the semantic approach to BI, and we love dbt, but every BI we tried made it feel like our business logic was scattered and repeated across tools.
Projects were hard to maintain, BI contracts were super expensive, and we felt like data tools just hadn’t caught up with the wave of more technical data people (who we now call “analytics engineers”).
Eventually, after trying every other option, we realised we had to build our own tool to solve these issues for data teams.
We started by “instantly turning your dbt project into a full stack BI platform” (you can see one of our early demos here!). We made sure everything was well-governed, kept in sync, and business logic was kept in one place: your dbt project.
We were building BI “The Lightdash Way”:
The Lightdash Way is about giving data consumers meaningful building blocks to answer their own data questions. Data teams maintain the set of metrics and their definitions, and everyone else self-serves, no SQL needed. - Hamzah, CEO
Just 2 months after our first commit to the Lightdash repo, we had users looking to do even more when developing in Lightdash. They wanted to be able to write custom SQL queries and use them to build charts, on the fly.
We get it. There will always be situations where someone needs something from the data team now and there’s just not enough time to write the SQL, add it to your dbt project, and sync it back to your BI tool.
We recognized that allowing our users to have the best of both worlds (the freedom of ad-hoc SQL exploration and also an easy path to sync everything back to their source of truth) was key to making Lightdash an even more powerful developer experience.
We wanted to avoid creating new logic in the visualization layer that should really live back in the metrics/dimension definitions. If it was going to be reused, we wanted to avoid creating another “source of truth” and scattering/repeating business logic across multiple places (which was a big reason why we started Lightdash in the first place).
Ultimately, we want to build the best BI tool for data teams. They wanted more power in their developer workflow. So, that’s what we’ve built today.
We’ve been building more speed and flexibility to the data developer workflow in a way that still fits into “The Lightdash Way” of building BI.
You can quickly write a custom SQL query and create a chart on-demand. Then, you can easily save these queries and write them back to dbt from Lightdash as governed, reusable models that everyone in your team can explore.
Data teams using Lightdash now have the speed and flexibility they’ve always wanted in data development. This seamless development workflow combined with our intuitive, SQL-free exploration means that your whole business will feel empowered by data to make better decisions, faster.
We’re really excited about this new way of building a governed BI with dbt. And we’re going to keep building more powerful, intuitive experiences that enable your entire organization to make data-informed decisions, The Lightdash Way.