Sigma + Databricks: A Great BI Tool for the Data Intelligence Platform
Introduction
Sigma is a powerful BI tool the team at SunnyData and I enjoy using for building dashboards. While not as well-known as Power BI or Tableau, and also while not bundled with Databricks like AI/BI Dashboards & Genie, Sigma brings a lot to the table, mainly by making building solid dashboards easy and enjoyable. Don’t just take my word for it. Check out this video below for a quick hands-on demo, plus a review below it.
Pros
Sigma gets many things just right:
The BI development experience is very smooth and intuitive. Most of the features that you can imagine you would need are probably available, and they are right where you would expect them to be.
Visualizations look fantastic and there is a good variety of charts/visualizations available.
Small learning curve means that as long as you know how to use SQL to retrieve the right data, you will find it relatively easy to start building.
Very flexible embedding capabilities.
The customer support and customer success teams are highly responsive, which can be a key differentiator for enterprises or startups looking for readily-available, personalized support.
Administrative interface is very clean and easy to navigate. I’ve rarely had to use the documentation for this type of work.
There are write-back capabilities available through what it calls “Input Tables”, allowing you not just to add read data, but add and/or edit data, live. These types of capabilities come in handy, especially for finance teams.
Supports connections to a variety of sources including Databricks, Redshift, Snowflake, BigQuery, and MySQL.
Neutral
Some of the points that fall down the middle:
There is no Git integration for version control. Sigma instead introduces what it calls “Tags” or “Version Tagging”. Not quite the same, but arguably an acceptable solution, allowing you to do additional development to an existing dashboard without affecting the production version until it is ready.
Sigma connects to the underlying data live. This is great from the perspective of not needing to orchestrate refreshing the BI layer and trying to keep it in sync with the underlying data. It is also great from a data security standpoint, but on the flip side, this means that even if the underlying data hasn’t changed, your compute will spin-up when using Sigma, every time, costing you $.
Cons
Like any tool, Sigma has its downsides. Some of the ones that stick out to us:
There is no free version of Sigma. The trial it offers only lasts 7-days, which I find to be a rather short amount of time, especially given the next point.
Cost-wise, Sigma is likely to be cost prohibitive for smaller businesses that are not tech centric, due steep pricing when user-count is low compared to some of the big players like Power BI & Tableau. On the flip-side, larger organizations or businesses hosting applications might find the pricing to be as favorable or more favorable than its top two competitors.
Historically, new Sigma features have lagged behind when it comes to the Databricks support. This is something I believe the Sigma team is working on improving, but worth noting.
Conclusion
We encourage you to check out Sigma paired up with Databricks for yourself or reach out to the team at SunnyData on how to get started with it. It truly is a fantastic BI tool that makes developing, deploying, using, and managing dashboards easy.
Josue is an Advisor & Databricks SME to SunnyData. For some more of his content, please follow him on in/JosueBogran and on other articles here at SunnyData.