ADOKI: Identical data on multiple analytical or data systems – automatic, in any data volume, efficient, fast, and secure!
12. 02. 2020
Reading time: 5 minutes
Augmented functions such as Artificial Intelligence, Machine Learning and natural language processing are becoming a key part of modern analytics platforms. And more. There are already tools on the market that free users from the previously necessary knowledge of data analysis applications. Now, they can simply use a search tool for queries, or even let the system preemptively provide answers that are of interest, though they have not thought to ask them.
In recent years, analytical tools have undergone a major revolution. At the beginning, there were mainframes; it took months to process queries. In the 90s, reporting tools appeared that allowed specialists to create reports on demand, which involved weeks to months of work. This solution is still often used today. After the year 2000, revolutionary visualization tools appeared on the market, bringing with them the benefit of modern graphics as well as simplified work with data and reporting itself thanks to “drag and drop” functionality. Even so, it turned out that several days of training and subsequent self-study were needed before someone was able to create a simple overview of the data. This is still not an optimal solution.
Obrázek 1: Není nutné znát SQL ani jiný skiptovací jazyk
Times have changed with the use of modern applications offering entertainment, information, and various services. This is increasing what is required of the operation and method of controlling analytical tools in companies. No one wants to spend a lot of time on the theory behind a reporting tool in order to understand how it works and use it sensibly. Imagine a situation where the analysis needs be carried out quickly, in a matter of minutes: “Why have sales fallen in a particular region?” In this stressful situation, the user, who rarely utilizes the reporting tool, has a hard time remembering how to create pretty visualizations, let alone how the advanced features work. In the end, he returns to the old, time-tested approach and builds his graph in Excel. It is also a fact that companies are not willing to invest in time-consuming training for their employees to become familiar with these tools and start using them actively.
Another obstacle is that nobody wants to wait a few days or perhaps weeks for an answer to their question. Often, they are not actually able to do so because if the answer comes several days later, it might happen that the information is already obsolete or irrelevant, and the company might lose a competitive advantage. This is why modern analytical applications change the approach to data and make it possible to put questions in the form of a search query in the same way we are used to in other applications. Thanks to its ease of use, the tool can be employed by a worker at any level, not just managers, as has been the case until now. All this without the need to know either the data model or how to control the analytical tool. The application is thus truly self-service.
Obrázek 2: Analýza dat formou chatovací aplikace
We are entering the next phase – as opposed to the current situation where people get information from the system based on defined requirements – where the system automatically proposes personalized content. The system is smart enough to offer the user additional interesting items and information when they are looking at a visualization based on their query. This is all a result of it continuously learning and recommending content based on end-user behavior – whether their profile, role, search history. It is similar to online shopping. If you look for a product, the system will immediately recommend appropriate add-ons, or what other people who bought that product purchased along with it. This is exactly the direction analytics is taking. The time has come for the system proactively to alert people to the information they should know and bring them a competitive advantage.
Published in IT Systems 12/2019, the author of this article Jan Bednář is a Al-Driven Analytics specialist at Adastra.