sublimd creates custom client dashboards with Cube semantic layer

Originally published at sublimd Blog

We are sublimd from Switzerland, an award-winning medical software platform to radically reduce paperwork in hospitals, enhance treatment quality and revolutionize clinical research.

In the beginning of 2019, we have received the first request for our new module sublimd Analytics from a client. At that time, we had established an open source business intelligence server solution in our product. We were struggling a lot with preconfiguring our analytics dashboards and ship it ready-to-use to our (non-technical) customers. Acting from necessity, we were looking for an easier solution to cover our needs: having full control over the customer configuration (tracked in a version-control system).

By chance, Cube attracted our attention on an independent comparison website. It felt almost too good to be true: Cube fits our technology stack including Node.js, MySQL and Redis. Which means a much simpler deployment than before. Beyond that, the framework lets developers the freedom to implement the frontend with your custom look and feel. After creating a small prototype, our goal of a seamless integration in our software suite appeared realistic again.

During the integration, we were really pleased how mature Cube already was. As our organization was de facto the first developer team who was working with the framework, we got an overwhelming premium support by the Cube team (Kudos to Pavel and Artyom). In return, we have sent back a lot of feedback and occasionally contributed to the project with commits to GitHub. It is our effort among other things to provide TypeScript definitions.

Our first version of the product still had the usual teething troubles. So we have improved it and a few weeks ago, the brand new version has been successfully rolled out to our first sublimd Analytics customer. The customer is happy, the software works like a charm and reveals unprecedented information of processes in the hospital.

What’s next? Our analytics module does not yet tap the full potential. We have thousands of health data points to pre-process and visualize in appealing charts. Cube is our framework of choice with building blocks including a caching and pre-aggregation layer to allow us to build a high-performance analytics tool for healthcare. This is another success story of open source software where a distributed team of engineers from all over the world solves a real-world problem.