Use of analytical data for development of IT services

How to use analytical data for development of qualitative IT services

Analytical technologies are widely used for information extraction from data to define sales pattern, to understand preferences of clients, to measure performance of business units, to adapt marketing strategies and infinite number of other tasks of business improvement.

Now, in 2020, after many years of support of different initiatives of corporate analytics, the increasing number of IT leading begin to understand that analytical technologies can be also used for improvement of their own transactions and services.

The analytics allows the IT operational commands to integrate the structured and unstructured data for the analysis, improvements or changes of quality of the provided IT services. It also helps IT commands to understand the major operational trends and to make decisions for optimization, correction and scaling of the services Bhanu Singh, the senior vice president for design and DevOps in "OpsRamp" (the company which is engaged in consultation of IT commands in the field of management and automation of hybrid environments

Until recently the analytics was generally used "post factum" when the IT organizations relied on old data for forecasting of future trends. "We had no adequate infrastructure, including computing power, for predictive analytics or the analysis in real time", - Singh explained. "Now the modern systems can integrate OLTP (on-line transaction processing) and OLAP (on-line analytical processing) that in real time to obtain information on functioning state and efficiency of business services"

Without analytics and careful monitoring of the IT organization work "on the fly".

If there are no basic data on how your IT transactions and services work, you will not be able to see when there is a problem, and, therefore, will not know what and where there are opportunities for improvement. It is important that engineers watched closely indicators that they could reveal and stop problems before they become aggravated and will affect users. Instruments of monitoring and analytics are key for different concerned parties in the organization Xiaohui of San (Xiaohui Sun), Technical lead of LinkedIn

To receive a uniform picture of the general behavior, use and trends, IT departments need to take operational data from appropriate sources.

"However, before IT specialists will be able to execute requests to these data, they need to be cleaned, normalized, enriched and indexed to provide necessary intelligence for decision making", - Singh warns

Applications

The analytics can give to IT specialists the deep understanding necessary for providing effective, fault-free services and transactions for corporate divisions of the enterprise.

At service of a computer system, such as Web server or cloud service, it is necessary to be ready to number of users which will use a system at different times. The analytics of the previous events can help to predict future. The analytics can help to define what loading will occur in unstable times. Benjamin Moseley, the associate professor of operations research from university business school of Tepper Universitet Carnegie – Mellon

For example, the retail website can usually have rather stable levels of traffic, but suddenly it experiences a heavy load of users in holidays and on sales.

If IT department uses a support service on the basis of requests, the analytics can be used for identification of trends in requests and offers of improvements and/or optimization technology and/or business processes for reduction of number of requests Igor Ikonnikov, senior analyst of IT department of InfoTech Research Group company

Network security is the one more IT area which can benefit by analytics. For example, on the website or in an internal system it is possible to trace users regarding statistically unusual behavior, - Moseley told. The analytics of last behavior can help to distinguish normal and unusual behavior

Monitoring

The system of analytical monitoring should be fast and rather powerful to process large volumes of the data generated with a high speed and a low delay that allows users to browse the latest data and indicators in real time noticed Xiaohui San

To implement this opportunity, LinkedIn uses the data warehouse open source which can service thousands of analytical requests per second. On the basis of it the command of LinkedIn created the platform of monitoring.

It is the platform of monitoring open source which integrates data from different sources, - explained Xiaohui San. We made it flexible in setup for work with different templates of metrics. Using different tools of the analysis and integrating the signals of an analytical system relating to specific areas on all business, a system allows LinkedIn "connect points in scale, giving a complete idea", he told

Necessary skills

For data analysis it is required that team members had an opportunity to formulate target questions and know what data types can be effectively available and provided to users using the corresponding software. It is also important to know how to interpret and apply results, Moseley emphasized. "For this purpose usually either considerable practical experience, or degree in data slicing pane or in the respective area is required"

There is a set of free modules and models which IT specialists can use for creation of the individual analytical solution.

But all of you equally should understand how these models / modules work. Igor Ikonnikov