Telekom Slovenije had been using analytical solutions for more than a decade but new data sources and demands from business users proved challenging for the existing analytics environment – the quantity and speed of data were simply too large and too fast to be handled with existing tools and solutions.
About the Case Study
Telekom Slovenije case study is describing the use of Exasol as a solution to provide most complex analytics.
Telekom Slovenije is a comprehensive communications service provider in Slovenia. It is recognised as the leader in the introduction and connection of new generations of mobile and fixed communication services, system integration and cloud services and multimedia content.
The Telekom Slovenije Group is one of the leading providers of communication services in the region of South-East Europe. In addition to the national telecommunication operator in Slovenia, it also operates through its subsidiaries on the markets of Kosovo, Bosnia and Herzegovina, Croatia, Serbia, Montenegro and Macedonia. The Group’s activities include fixed and mobile communications services, digital content and services, multimedia services and digital advertising, system integration and cloud computing services, construction and maintenance of telecommunications networks, and conservation of natural and cultural heritage in the Sečovlje Saltpans Regional Park. It operates the most reliable and high-quality telecommunications network in Slovenia, as well as one of the most complex backbone networks in the region of SE Europe.
Challenge
Telekom Slovenije had been using analytical solutions for more than a decade. New data sources and demands from business users proved challenging for the existing analytics environment. The quantity and speed of data were simply too large and too fast to be handled with existing tools and solutions. “Our biggest challenge was the speed of carrying out queries over large data volumes and the discontinuation of the data intelligence department which dealt with the analysis of more complex data”, remembered Borut Rožac, software engineer at Telekom Slovenije.
The company decided on centralized processing of various data derived from its numerous services. As analytics were handled by business intelligence department new challenges arose from the side of data scientists which wanted to perform highly demanding queries on large data sets. “Our former analytical solutions simply could not handle complex queries over huge amounts of data – we had over a billion and a half lines of data connected to a large number of other data sources”, said Rožac.
Solution
In 2017 Telekom Slovenije decided to thoroughly renew its analytics infrastructure. A goal was to find and implement a scalable and therefore future-proof solution. The IT department and BI business unit set out the use case for some comparative analytical solutions. The aim was to review different analytical solutions against each other on the same virtualised hardware platform and faced them with one billion lines of data and complex queries.
We also compared the ability to use analytical tools like Python and R directly on the database. We checked the support for different programming languages, and the possibility of integrating the database through the ORM-functionality. Finally, we compared the costs aspects: cost of licenses, implementation and maintenance, coupled with the required level of system administration.
We were blown away by the exceptional performance and responsiveness of Exasol. And this was the only database we have worked with that required virtually no administration once it was set up.
Borut Rožac
In the end there was one clear winner: Exasol in-memory database.
Benefits
Exasol in-memory database environment is now used by Telekom Slovenije for advanced analytics and machine learning tasks. “We were blown away by the exceptional performance and responsiveness of Exasol. This was the only database we have worked with, that required virtually no administration once it was set up”, said Rožac.
Exasol database enabled us to efficiently process vast amounts of data from different sources which simply hasn’t been possible of viable before.
Borut Rožac
The Exasol analytical database more than fulfilled the expectations regarding the speed of advanced querying. This was demonstrated with by far the fastest generation of table extracts for analytical software Tableau that is used by data scientists and business users at Telekom Slovenije and derives insights that are presented via personalized dashboards.
“Exasol database enabled us to efficiently process vast amounts of data from different sources which simply hasn’t been possible of viable before,” concluded Rožac.
Key benefits:
- Performance
- Ease of use
- Compatibility
- Value
Implementation
The implementation of Exasol in-memory database started early in 2017. With the help of CRMT experts Telekom Slovenije implemented the solution in one single day. Initial load of one billion records was finished within the week though it started data collection from the “flat” files. CRMT also carried successful integration of Exasol with our big data environment. They integrated also other applications – Tableau, Phython and R what enabled Telekom Slovenije to take maximum advantage of their current infrastructure.
“Everything went smooth, the implementation was very easy and quick and the migration from the virtual environment to the physical machine, used in production was equally fast”, Rožac noted.
Conclusion
Today Telekom Slovenije is using Exasol in-memory database as its main platform for the independent work of data scientists. Business users enjoy excellent system responsiveness. System administrators praise its scalability and ease of integration with other existing hardware and software solutions and data sources.
With clear monetization of data in sight, the company plans to expand the use of Exasol throughout the companies in Telekom Slovenije Group.
You can also watch the presentation of Borut Rožac, Telekom Slovenije on Exasol Xperience 2018