Ömer: Please describe general developments in the technology and how your company will contribute to the field.

Osman Demir: We are working on data analysis and database systems, and in particular, graph databases and their visualization. In the realm of databases, there is this classical approach, in which data is stored in rows and columns, in other words, in tables. This is how most of the companies in operation today store their data. While this may seem as a good way for data storage, and in some aspects it really is, it is not a very suitable method for certain operations, such as relational searches. Relational searches are type of searches in which we search for individual data through its relations with other data. Such a search is applicable in tabular databases, but as the number of relation chains increases between datas, this becomes inefficient very quickly. Thus a need for graph database emerges. In this database, data is stored as nodes and their connection with other nodes, called as edges. In such a database, relational search operations become much more efficient. However, companies do not have this type of database, they have tabular ones. So we integrate graph database in their classical databases, so that they can benefit from the advantages of graph databases. Also, we develop a visualization tool and query language for the graph databases, so they become usable. But we also provide various data solutions to institutions if we are demanded.

Onur: What are the main applications for this technology?

Osman Demir: There are many applications, since data is everywhere nowadays. We have worked with companies, institutions, ministries, and they all wanted to have the graph database in their arsenal as well as their classical databases, so they can analyse and get results from their data as efficiently as possible. For example, one project we have been working on lately is for a ministry. They wanted us to provide them with data analysis tools to determine the efficiency of their employees, and we developed said data analysis tools for them. It not only uses graph database, but also machine learning in its analysis. Now, by analyzing the work data of the employees, the efficiency of workers can be seen by the managers, using this tool. I hope it will be useful for them to improve their operations, so they can serve better to our citizens.

Ömer: What are the risks and benefits of this technology?

Osman Demir: There are risks associated with data business of course. The confidentiality of data is often a headache for our business. We are, in general, not provided with connection to the data of institutions from outside, so we have to do our work at institutions in general. Also, we have to make sure that we address the privacy concerns of our customers through our development procedures. Also, there are the ethical issues. We need to make sure that our data analysis tools are accurate, they need to provide answers with a level of certainty. They need to be transparent, we need to be able to understand how our tools provide us with results if there happens to be the need. Furthermore, we need our data technologies to provide unbiased results, that is, they need to be fair. Machine learning applications in data analysis is a particular example on this. If the initial data provided to the machine is biased, then the machine will continue to produce biased results if proper precautions are not taken, and this can be a major problem. For example, in our work efficiency analysis technology we have mentioned before, inaccurate analysis results can cause workers to be depromoted, or even lose their job, which would be very unfair for them. Therefore, we are trying to develop our technologies with these concerns in mind.

Onur: What sort of modifications will be made to the technology in the future?

Osman Demir: Data analysis is not going to be insignificant anytime soon, since it is such an integral part of our lives now. But the way we perform these analysis has much room for improvement. More efficient database structures and analysis algorithms can be developed, and integration of machine learning in data technologies can be increased further. We are also working on these developments in our company, so we hope to become a significant actor in data technologies in the future as well.