How to build analytics platforms

Reading time: approx. 2min

Part 3: Individualized workflows and dashboards

The more information a platform represents and the more detailed the individual analytics projects are, the stricter is often the way in which projects must be planned and implemented. But what would happen if the individual project steps were based on the company and individual dashboards could be reused? That is exactly what we are looking at this time.

No added value without individualization of the workflow!

Input masks, sub-steps and checkboxes – everyone knows them and has to deal with them in various environments. So that analytical environments can perform their work, they are inflexible in their operation and often rigid in their sequence.

If you must get used to a new solution and find your way in it, it can take some time until corresponding data projects can start and deliver results. At the same time, it is time-consuming to adapt existing processes to the new environment. This is one of the reasons why companies often find it difficult to replace existing platforms with new ones. Nevertheless, processes and projects, especially in the area of analytics, have often grown in scope and complexity just as much as the companies themselves. A rigid user guidance is therefore an obstacle.

An approach on how to solve this: Customizable workflows that can be freely configured in their sequence and operation. In this way, existing steps and projects can be adopted and, at the same time, new functionalities can be added and streamlined. Transferred and adapted, the individual steps also serve to improve documentation and transparency of the entire process.  By customizing and compiling the individual project and process steps, prototypes of individual use cases and projects can be created and tested more easily and quickly in the platform – this point is becoming increasingly important for companies, not least in terms of scalability.  But only a few platforms already offer this possibility.

Data Science Workflow

CONCLUSION: With the possibility of compiling the individual project steps yourself, you get an analytical environment with which companies can create and manage their own data projects without long training periods.

Reusability and repurpose instead of disposable dashboards

What applies to the process steps in the above point also applies to the creation of dashboards or views. In addition, a flexible use of analytical environments requires a free display of information. You need to give each participant or user group the ability to freely set up their views or dashboards, or to customize them according to their needs. Not every group needs all or the same information. If you can configure the views yourself, you avoid an overload of information and promote more effective work. What is the use of all the information if the overview is lost? Individual widgets can be used to assemble targeted project views that display and edit exactly the information that each participant really needs.

But how can this be implemented easily? This shows how fundamental an intelligent role concept is! If you can control authorizations at a component level, you can quickly implement individual dashboards. There is no need to create individual views for each dedicated user group – instead, a default is set up and the individual elements would be shown or hidden, depending on the authorization. And how do you distribute the dashboards you create to users? A simple implementation would be if dashboards could be shared with other users via a simple link.

Analytikplattform Data Science Plattform Dashboards

To put it simply, this can be imagined as a kind of “template”, as it is used in content management systems like WordPress. Thus, several parties benefit directly from the creation, which reduces the workload. At the same time, these dashboards can also be assigned to OTHER projects. Advantage: Faster progress in individual projects and easier creation of new projects. In this way, views can be distributed throughout the entire company via a deep link in the form of a URL. This factor is essential when it comes to the scalability of data projects.

CONCLUSION: As with workflow, individual views are important if analytics environments are to be used as widely as possible. With the combination of an authorization concept and the possibility to distribute, the workload and training period can be reduced.

Are you looking for an innovative data science platform? Then discover YUNA!


In the next part: The scalable connection of data sources and the possibility to create new data products and business models. Read now!