Hi,
I have written something but I was unable to upload... or I don't understand the process to proceed. Let me share it:
TASKS AUTOMATIZATION IN THE DATA CYCLE
We are working with clients from all over the world, focusing on those who speak Spanish. In all the approaches we have had with our clients, the primary needs were to solve different business analytical needs. The themes and cases may vary, but basically, all clients started with an analytics guideline focused on the analytics objects they should get. Depending on the degree of maturity, migrate from an Excel-based way of working to apply predictive analytics to your models in your cloud data warehouses.
Data processes generally do not appear anywhere on the initial roadmap. However, the automation of data processes, the integration with business processes, and as a more ambitious case, turning Sisense into the core of decision making by applying actionable analytics, are usually not in the first phase. However, these automation processes are, ultimately, the core of the analytical solution. From being able to build cubes dynamically based on data needs, being able to communicate information obtained from Sisense to other processes, or invoking business processes from a dashboard, after all, it is what we call Maslow's pyramid of data, which ultimately makes the analytical solution a determining factor in decision making.
To achieve this, the intensive use of the REST API in conjunction with components based on BloX, has been shown to make a reality that, being an open platform, I can integrate it into the data governance of organizations, and, in many cases, boost it. In the case of OnPrem clients, the use of Linux shell scripts together with crontabs is vital, but it is not exclusive, since we take advantage of the Jupyter Notebooks environment to be able to perform processing logic within the cubes, giving the power, not only to generate new data sets, but also to orchestrate and generate calls to business processes through Python.
Think of Custom Code for what it is, code to enrich the data.
From being able to manage security by registering or deregistering users, to generating PDFs based on existing dashboards and, once generated, connecting with a messaging service for distribution throughout the company. In many cases, they can also fill certain gaps in standard SQL. If you have worked with Oracle databases, for example, to obtain differences in data, you are used to using the MINUS statement above NOT IN. This is not available in standard SQL, but you can fill that gap through Python by calculating the differences of data sets optimally.
