🌈 Update: We built Rakam UI from scratch

We have been working hard for the last 6 months to build a great team and develop a newer and better version of our app. The first version served our purpose well for 1.5 years but we started to have difficulties adding new features, on-boarding new team members and adopting new user base because of the design.

The first version of Rakam UI is developed with AngularJS and the UX was not optimized enough since we didn’t actually know our potential user base before developing the product. Over time, we had 3k users, both small and medium-size enterprise companies so we analyzed which features are being used, what are main pain points and why we couldn’t be able to convert more leads.

It turned out that medium size to enterprise companies already have a data pipeline and providing Rakam platform as an end-to-end solution doesn’t work for everyone. Especially since the enterprise companies collect all of their company data (product, marketing, financial etc.) into their data lake and use BI solutions in order to analyze their data. What we had realized is that these BI solutions are usually not useful enough for product teams so they still prefer using SaaS analytics solutions such as Mixpanel and Amplitude. This is mainly because most BI solutions try to be generic, you need to spend time to adopt it for your use-case and even than it’s still not perfect, product teams usually depend on the data team to build up the user behavior (funnel, retention etc.) reports, we try to remove this bottleneck so that you would have more time, you can ask more questions and dig deeper so that you can have a better understanding of your user data.

However; this option is not possible for some enterprise companies since it requires sharing customer data with a third-party company. Also even if they can still use these SaaS solutions, it’s not possible to see the big picture since they don’t have access to company data other than the product data. When you have more than hundreds of event types, it’s often challenging to use these SaaS tools since they’re designed for medium-size companies so we decided to provide Mixpanel-like interface (well, much more than that for enterprise companies)to companies which have their own data-warehouse.

Also, we had some UX issues which caused the on-boarding flow to take too much time so we decided to re-design the app and turn it into a separate product which can work on existing data-warehouses. We used Kotlin in backend chose Vue.js as our front-end framework. The final solution is more like Looker for product data, you enter your database credentials, map your data and start using it without any need of writing SQL. When our high-level reporting features are not able to cover your use-case, you can use our SQL reporting in order to create custom reports.

Here are some of the features that we added to the new system:

  1. Ability to add label, category, and description to tables and columns with taxonomy feature. It’s especially useful if you have hundreds of tables or columns.
  2. Advanced team features for enterprise companies. (table, row and column level permissions, auditing and feature flags)
  3. Ability to create custom SQL reports with variables. You simply embed variables, parametrize SQL queries and create custom reports with custom UI without any need of writing HTML or CSS.
  4. Ability to create OLAP cubes (summary tables) from your raw event-data. In the segmentation feature, you simply run your ad-hoc queries and we take of caching, incremental materialization for you.

The new version is live at https://app.rakam.io feel free to try it out, give us feedback so that we can make it even better!

Author

Burak Emre Kabakci