data data-revolución!

In 1700’s there was a problem in tin and coal mines; flooding. This slowed down the production of coal and tin, so people with money said that they needed to do something about it, they got Thomas Newcomen, and he was assigned with the task of improve the way they pump out water from places, by 1705 he came up with the Beam engine (the first iteration of the steam powered engine). So for me, this is where all the technological revolution started, forget fire, forget the wheel, the beam engine was doing the work of people by pumping out water from coal and tin mines. This was the first step on industrial revolution, the next one was taken after the Beam engine was improved by James Watt (hence the Watt unit to measure power).

Now ~300 years later, data, and the way we look at it, has started a mini-revolution. Jobs, sciences, careers, and people lives have been changing because of it. This will only continue to grow, as data is generated every second, but there is so little time to get value out of it.

OK, so enough of history. Due to this I have been playing and experimenting with different data sources, and ways to get value out of it. And because of this, I tried some “data analytics” tools. There are many, but I will try just to mention a few projects that look interesting for anyone to check them.

  • Tableau, right now tableau is on the top ones, why? just because it is simple, and easy to use. I can create a read only user for someone and have them their way without worrying about them doing something weird to the database.
  • Grafana, for time series data this is perfect! It the second one I most use, plugins and all are very good, but they require a bit more of data massaging than tableau, and a bit more experience on the SQL side, so that you can convert your data into time series or at least something meaningful.
  • Kibana, perfect for seamless integration with your elastic search data, it requires a lot of workarounds and effort to plug it to your custom db, and also requires a bit more technical skills to improve the dashboards.
  • Metabase, something opensource that wants to step into simplicity of tableau but still the UI and its dashboard configuration has some polishing to do, but very straight forward on its deployment, (1 command to run and test it using docker)
  • Superset, still in incubator mode, so right now “use it on your own risk” type of project. Looks very nice and customization, but it requires some effort to plug it to my common data sources (SQL Server, PostgreSQL, MySQL).
  • Looker, I just found this looking into metabase. But honestly I don’t know anything about this other that it is not opensource, and you have to pay $$.

Hope, that if you are looking for options, the ones above can help you make a decision or to guide you to try them.

el_avena off…