Technoserv of DSW - Data Science Workspace

2018: Data Science Workspace service

On December 7, 2018 Technoserv announced creation within the competence center of Big Data (CCBD) of division for development of Data Science Workspace service – a uniform working space for analysts by data.

Data Science for December, 2018 is one of the most perspective and fast-growing IT directions. Specialists investigate and develop forecast models of machine learning which help business to increase income, for example, due to preparation of personal offers, or to reduce costs, calculating the probability of accidents and defects on production. Because of relative novelty of this direction to the companies it is difficult to define tools, sufficient for work, to develop regulations of work of divisions, to organize control. With the need for special infrastructure and support from engineers of Data Science-специалисты often use the software for work.

To minimize the probability of data loss, documents or results of a research of specialists, the Technoserv Cloud command and competence center of Big Data of Technoserv company developed the concept of a cloud service for simplification and acceleration of work of the Data Science command. Recently created TsKBD product direction which was headed by Timur Mishin is engaged in its implementation.

Data Science Workspace (DSW) service of the company will be able to receive on a subscription both based on a cloud the Technoserv Cloud, and in the private cloud. The service offers the environment for joint work of the DS command based on Jupyter software stack with an opportunity to select and manage profiles of computing resources and a shared folder for data and researches. In the near future adding of following features is planned:

• management system for computing user profiles (one calculation on one server with complete or partial use of resources);
• system of accounting of resources and budget management of calculations;
• monitoring of activity of sessions and projects of users and overall performance;
• integration with GitLab with an opportunity to save all researches and their intermediate versions in the general repository and to receive the previous version of researches a one click;
• configurator of the preset libraries ML and their versions;
• CPU and GPU of calculation on demand with payment of only actually spent resources;
• cluster computings on Spark on demand with payment of only actually spent resources.

In case of the choice of a cloud the Technoserv Cloud which has a full range of certification in the field of security of information systems and security of personal information data will be stored and be processed in the territory of Russia.

The world trend indicates that similar services have an excellent future in which the companies with a world name will participate. At the same time abroad already actively use cloud services, despite rather strict requirements of GDPR – regulations on personal data protection. Georgy Shatirov, director of competence center of Big Data of Technoserv company

When developing a prototype we analyzed experience of foreign competitors, conducted an interview with representatives of business and IT of the different industries. The analysis showed that for preparation of a simple and convenient product in use we will need completion of a part of available open-source-solutions, including deep immersion in Kubernetes and Docker for virtualization of calculations. The direction and the used technologies are very interesting and demanded, and we invite in our startup both professionals of the case, and the burning enthusiasts in the direction of DevOps, Full - stack - development, the product analysis. Timur Mishin, product area director of competence center of Big Data of Technoserv company