How to make invisible data insightful? Tristan (NoCode-x) sharing his experience.
It all started with the goal of analyzing data in the blind. Our user is working for an organization with the highest security standards, but they have no access to data scientists with coding skills. Tristan, CTO of NoCode-X, enhanced the Virtual Data Lake by Roseman Labs with an application built with the NoCode-X platform. Our user, wanted to build a platform that enables secure aggregation of cyber data, particularly adherence to best-practices. NoCode-X is the development partner of our user. With NoCode-X, you can build secure web applications without developer skills which is accessible and affordable for all. The goal of Tristan and the user is to create interactive dashboards from the Virtual Data Lake output.
In this article, we discuss Tristan's challenges and accomplishments in creating a monitoring system to make data understandable for everyone - and in this case, for our user. As a bonus, Tristan explains how you can visualize data from the Virtual Data Lake without writing a single line of code.
The Virtual Data Lake by Roseman Labs enables organizations to collaborate on sensitive data without exposing the source data. You are in full control of your data, and you can analyze details as if it was a single source, ensuring controls and protecting against malicious insiders. However, not everybody has the luxury of data scientists in the team. According to Tristan, this is not required to gain valuable insights on your data sets using VDL. The offered workflow is to use the Crandas Python package (a tongue-in-cheek reference to Crypto Pandas), an equivalent of the well-known Pandas package but specially designed for the Virtual Data Lake. When using the Virtual Data Lake, you will get access to examples, tutorials and documentation. These inputs, and the support of the Roseman Labs team resulted in the final product – a dashboard built with No Code X.
Our user is using the Virtual Data Lake to ensure the data privacy of individuals. He found Tristan and explained his problem. The user has multiple data sets, but they all contain sensitive privacy information. When using the VDL, he can benchmark salaries for different roles within security without disclosing individual salary information of respondents. The biggest challenge for Tristan is to connect the VDL with NoCode-X platform and set up a dashboard in which the user can analyze the data and filter on the properties. In the current landscape of data visualization, the options are growing (example: Microsoft PowerBI). Tristan is looking for a scalable solution which is also applicable to other future challenges.
Tristan started with the overview - what is the information that is needed and how can he visualize that. In this overview, it's important to understand your data characteristics – which properties do you have available and how are they currently mapped. If needed, you might want to map (or remap) some data to make sure you have clean insights. Within NoCode-X, there are a number of visualization options available such as line chart, bar chart, checkboxes, donut charts, pie charts, etc. In this case, Tristan uses the pie chart and line chart to visualize, and checkboxes to enable filtering. With an easy drag-and-drop interface, he is creating the outline of the dashboard. Tristan was excited to see how easy and frictionless the process was.
In less than a day the connection between the Virtual Data Lake output and NoCode-X was created. Tristan: “I never imagined this was going to be such an easy setup.” The output of the Virtual Data Lake is available through the Crandas Python package, which he connected with the NoCode-X application – using a small python script. You can build the script based on the documentation provided by Roseman Labs, available within the Virtual Data Lake. The script, running as a serverless function, is used to automatically fetch data from the VDL and expose this data by means of a REST API to the NoCode-X application platform. The biggest challenge, if any, was to keep the connection running and make sure the dashboard is fetching the data after the user interacts with the dashboard.
To make the data visual inside the current implementation, the serverless REST API and NoCode-X components are used. The impact of this implementation is that it can feel like the dashboard needs to “wake up” and you might need to wait a few seconds for the dashboard to respond. To overcome this in the future, you can try out different implementations such as for example not choosing for a serverless setup.
So what's next? In this case, the NoCode-X application will be used for multiple projects similar to this proof of concept. This project was enjoyable and satisfying for all stakeholders, at an incredible fast pace. In future projects, Tristan is thinking of extending the current functionalities of his dashboard. To name a few of these functionalities – a save functionality which allows you to save a certain set of filters, combining date from multiple data sources or using the output of the dashboard as input for another dashboard or even business process.
The conclusion: if you are looking for an easier way to analyze your data in the blind or if you don't have any data scientist in your team, give No Code X a try.