Can you still imagine navigating through the city without any aids? What do people who seem to do this effortlessly do differently? The key lies in an approximate inner city map that successful navigators use. In this map, they pay particular attention to the cardinal points and can thus relate their starting point and their destination on their inner map.
People who find it difficult to find their way, on the other hand, often just try to memorize a sequence of turns and specific street names. If they forget a turn or a street is closed, their strategy collapses like a house of cards.

The situation is similar for efficient and successful agile data analysis and data management: real information depends on its context and transparency. The ability to make information visible and to recognize and react to changes at an early stage is very helpful. Facts without context have less added value for us as humans; we find it harder to remember and use them.

The visualization of data is still neglected in many companies. Conventional architectures with complicated data models that rely on storing and linking different data tables rarely combine this with consistent and continuous visualization of the data.
In Germany, a digitally developing country, it is still common practice in many companies to extract data from databases and then visualize it in Excel. As a result, the overall data is not very transparent and comprehensible for experts without an IT background. There is also a major obstacle to making it truly visible and using it regularly. For many business decision-makers , the data, its links and its actual potential remain hidden.

Contrary to the widespread practice of viewing data visualization as the last step in the process, both data scientists and customers benefit from visualizing data at the beginning and during the course of a project. This is because dynamic visualizations give us as humans a much better insight into the data we are using. Especially via dashboards with graphics that can be adjusted and changed interactively with a click. This not only means a better overview of important properties of the data – working hypotheses can also be quickly checked, rejected and replaced by a better hypothesis.

Our core strategy at moresophy is to map customer data of all types and from various sources in dynamic diagrams at an early stage and to visualize this data during each work step. This means that we already use them during data cleansing and data mapping in order to be able to optimize the respective step at an early stage. We want to know what we are doing and continuously review our own ideas and approaches.

This creates maximum transparency for customers and specialist users as well as for developers – even before we deliver the end product. The flexible architecture of Visual Analytics also enables the integration of third-party sources, whose possibilities and limitations often only become apparent through systematic visualization.

Case Study:  Pre-Sorting of Damage Reports from an Insurance Company

An insurance company would like to automatically pre-sort incoming damage reports and in this way reduce the amount of work involved in accumulation damage events. Accumulation losses are losses caused at a similar time by the same loss event, such as a storm or flood. The challenge for insurance companies is that a large number of claims are received at the same time and claims handlers cannot keep up with their work. At the same time, the staffing market is empty, as accumulation loss events usually affect many insurers at the same time.
The aim of appropriate software is therefore to support claim handlers in their work and thus enable faster approval of actual accumulation losses.

How can it be ensured that an applicant was actually affected by such an accumulation event? Weather databases, for example, are available as possible third-party sources. By continuously visually checking which third-party sources prove useful, better decisions can be made during the work process. One advantage of this is that unnecessary work and therefore unnecessary costs are saved because it is possible to see at a very early stage to what extent additional data contributes to improving the data quality and also the confidence of AI models. In agile, cyclical rapid prototyping, it is possible to work with the addition of external data and test the achievable gain in quality and significance of the analysis. The time and energy gained can then be used to fully integrate the third-party sources that are deemed reliable.  

At moresophy, we use visualizations continuously from the very beginning of the work process in order to constantly review our own path and initiate clever changes of direction at an early stage. Like the inner map, which helps with navigation and is often more error-tolerant if a turn-off is not available as planned.