CTI and ER-Telecom develop a self-service system based on a dialog bot

2021: Integration of Robotic Assistants with the Proactive Service System

CTI and ER-Telecom have completed the next stage of the introduction of a robotic intelligent self-service system based on the contact center of ER-Telecom Holding JSC. This was announced by CTI on January 10, 2022.

The importance of the ER-Telecom contact center has especially increased in the past two years, when, due to the restrictions associated with the pandemic, the remote service format has become almost the only possible.

The platform-based dialogue bot omnikanalny Omilia Diamant allowed contact center operators to cope with the increased load. The redistribution of routine processes digital assistants through has freed up resources to address more complex issues. Thanks to this, the speed of service has increased significantly and labor costs have decreased.

As of January 2022, the dialogue bot supports the processing of more than 30 topics without the participation of contact center operators. In order to maximize the effect of automation, at the beginning of the project, the most frequent requests were analyzed, which were processed by contact center operators. As expected, a very large stream of calls was associated with repeated, routine requests regarding the current account balance, problems entering the personal account, and television settings. These questions were the first scenarios that were automated in the project. The introduction of automated scenarios was accompanied by constant monitoring of customer satisfaction.

Overcoming barriers in customers' perception of the dialog bot has become one of the main tasks of the project. Thanks to a careful study of logic and scenarios in one year alone, it was possible to increase customer satisfaction with the work of the robot from 4.3 to 4.8 points on the five-point CSI (Customer Satisfaction Index) scale.

The company achieved this high score by redesigning use cases and speech modules, bringing them to unambiguously interpreted wording. For example, a large jump in positive ratings occurred after the introduction of the "send contract number by SMS" scenario, since the talk of numbers and hearing perception is problematic, and obtaining such information in text form allows the client to use this information for various purposes.

In the next phase of the automation project, more complex scenarios involving the absence of the Internet, setting up tariff plans, monitoring equipment, their remote diagnostics and rebooting were subjected.

The introduction of the dialog bot required a significant processing of scenarios, the logic of which is fundamentally different from button IVR and text bot, "said Petr Smirnov, head of the automation of contact centers, ER-Telecom. - We analyzed hundreds of calls and concluded that the client addressing the text bot is aimed at solving technical issues, while in the voice bot the basic call is, for example, a balance issue. Thus, the preliminary analysis of client preferences, carried out by us before the implementation, ensured 80% of the success of the project.

At this stage of the project, the integration of robotic assistants with its own Proactive Service System (SPAS) was implemented. The system online evaluates the quality of services and finds malfunctions at the network node, at the entrance or in the apartment, assigning certain triggers to the problem. When a problem message is received from the customer and the relevant trigger labels exist, the customer is automatically sent by the bot to the technical support of the relevant topic. This scenario allows you to create service calls for solving technical problems without the participation of the operator, which speeds up the solution of the customer question and increases the level of satisfaction with the service.

At the same time, scenarios were introduced for proactive equipment monitoring, digital television connection and the ability to remotely reboot the equipment when the client reports technical problems.

In any of the scenarios, the bot itself clarifies whether there are any questions left, and offers the possibility of transferring the second support line to the operator. This is also the case when the bot understands that this topic is not yet automated.

As of January 2022, 36% of requests to the ER-Telecom contact center are resolved without the participation of the operator. This value includes the total automation efficiency of both the push button IVR and the dialog bot.

{{quote 'The main result of the project for us was the movement of focus in automation on how customers perceive this automation, "said Ksenia Zueva, project manager for customer service automation, ER-Telecom. - Attempts to maximize operator replacement with automation equipment would ultimately kill all loyalty of our customers. In the process of implementing the project, we realized that it is necessary to abandon the idea of ​ ​ austerity in the workforce and adopt a new paradigm for the development of the company - a client-centric strategy. }}

ER-Telecom is interested in scaling the voice IVR project for B2B tasks, and there are already several cases from internal customers - these are secretaries and a technical unit.

The project for the introduction of a self-service system based on a dialog bot in ER-Telecom is in full-fledged industrial operation mode with the active development of additional functions. Interest in using the system led to a significant increase in competencies within the team and allowed to build a coordinated system of interaction. Colleagues solve most of the tasks on their own, and we, in turn, provide assistance on request, "added Leonid Perminov, head of the Contact Centers, CTI. - The project showed that with the help of bots you can not only reduce costs, but, more importantly, reallocate resources, transfer the solution of routine issues to robots, and free operators to interact with the client on complex issues.

2020: Completion of AI-based robotic self-service system

On May 20, 2020, it became known that the business integrator CTI completed the main stage of introducing a robotic self-service system based on artificial intelligence Contact Center telecommunication the operator "." ER-Telecom Now robots successfully respond to customer requests in both voice and text channels, helping almost two thousand contact center operators process more than a million applications monthly. ER-Telecom customers positively appreciated the new capabilities - 47% of calls to the contact center are solved without human participation after one month of system operation in productive mode.

Faced with an increase in the number of calls to the contact center, the leadership of ER-Telecom decided to follow the path of automation, without increasing the number of incoming lines and not inflating the staff of operators. Automation also addressed the costly issue of hiring and training a large number of personnel related to the flow of operators (especially on the first line), and the uneven distribution of load on contact center specialists.

As a result of the competition to select a virtual operator platform based on artificial intelligence technologies, ER-Telecom focused on the decision of Omilia, which has proven itself well in the market of Russia and the CIS countries. The solution met the technical requirements and had the best price-quality ratio. Omilia's omnichannel platform was suitable for both text and voice channels, although competitions were held separately so as not to limit competition. The common solution found optimized the costs of implementing the system, reduced the cost of purchase, ownership and maintenance.

The CTI integrator was chosen as the general contractor for the implementation of the entire turnkey work. The complexity of the project was to integrate artificial intelligence into the existing contact center, the voice and text channels of which are implemented on technological platforms of different manufacturers. It should be noted that with the Genesys platform, the Omilia solution was able to integrate through open APIs for bots, which made it possible not to violate the integrity of the architecture and maintain the reliability of the system as a whole.

For the first stage of automation, the simplest scenarios for interacting with ER-Telecom customers were chosen, which make up 50% of all traffic: "find out the balance," "how to make the promised payment," "how the service center works" and other consulting simple requests. Next, the scenario of checking the technical capability of connecting a new service in relation to the address of the future client was automated. The bot analyzes the client's request, checks the technical connectivity at the address indicated by the subscriber and immediately transfers it to the second support line to issue an application, bypassing the first. The issue is resolved quickly and professionally, which positively affects the client experience of ER-Telecom subscribers.

Creating scenarios for a self-service system is a complex process that requires careful study. It is important to combine the efforts of employees within the organization, experienced integrator engineers, as well as involve focus groups in the process and further carefully test everything. The logic is very different from button IVR, it is complex and has variability, each step is divided into many subtasks, and only by experimental means can you find out how a real person speaks. The process itself affects both internal information systems and end-to-end business processes in the organization, which also need to be put in order. All this is done so that the result will bring tangible benefits to the business. Therefore, the claims that you can introduce a full-fledged bot in a couple of weeks and get the maximum effect are a little exaggerated, - comments Leonid Perminov, Head of Contact Centers, CTI

One of the tasks of the project was to reduce the load on the first line due to intelligent routing. Now the virtual assistant serves more than 30 topics on its own. For those topics that are not yet fully automated, the robot determines the client's intentions and immediately transfers to the desired group of second-line operators.

In the near future, a scenario will be launched that adds proactive equipment monitoring to intelligent routing: if the client reports that the Internet is not working, the system can see that the problem is a hardware malfunction and immediately redirects the request to the second line for diagnostics. You can also remotely reboot the client system using proactive diagnostics.

The artificial intelligence of the system can support a free dialogue mode with the client, allowing you to solve several problems in one session.

The Omilia DiaMant platform has already firmly established itself in the global market among solutions for building omnichannel Conversational AI portals. Our solution works in more than 30 contact centers around the world and all our customers show impressive business results in their projects. This is also because our partners make a very big contribution to the success of the projects. Specifically, the ER-Telecom project brought together an excellent team of like-minded people, both from the client and a partner - the Russian integrator CTI. This team was able to quickly master all the advantages of our platform and achieve, I am not afraid to say - one of the most impressive results of the introduction of an intelligent virtual assistant in Russia, - says Platon Begun, Director of Business Development of Omilia in Russia

It was planned that the project will pay off in 2-3 years, but I think we will reduce this figure. The target business indicator that we wanted to achieve by the end of 2020 is the processing of 50% of our customers' requests using voice and chat bots. Completely without involving operators. Without compromising the quality of service. Already in April, after 1 month of working in productive mode, we received 47%. The average running time of the self-service scenario is also significantly reduced: the request in the system is processed in 17 seconds instead of 40 previously scheduled. This advance allows us to now say that by the end of 2020, robots will handle 60-65% of all call traffic. Success here determines both the right choice of the Omilia technology solution, and the coordinated collaboration of a team of ER-Telecom employees and CTI engineers. We understand each other with a half word and do not ask again, which allows not only to advance the schedule of work, but also to include additional scenarios, - commented by Mikhail Zykov, head of client service automation projects at ER-Telecom

But the struggle for customer loyalty moves ER-Telecom to improve service and the next stage is planned to automate requests with anticipation of customer expectations. This means that artificial intelligence will access internal information systems, receive information about the client from them and, as a result, most comprehensively respond to its request. For example, when a client requests the availability of an Internet service, the system will automatically check whether there is money in the account, and depending on this, will give recommendations.

2014: Complete Cisco Platform Contact Center Implementation

On April 2, 2014, CTI announced the completion of work on the implementation of a contact center on the Cisco Unified Contact Center Enterprise Edition platform for the Russian telecommunications holding "Дом.ru."

Project Tasks

With the help of specialists CTI "Дом.ru" , I got an advanced contact center that allows you to dynamically route incoming calls and calls, get up-to-date and accurate statistics on each client request. The contact center includes a call recording and call quality assessment (QM) system on the platform. Verint This project is unique in terms of technology. Finesse Cisco , based on Web 2.0 technology and designed for the simultaneous operation of more than a thousand contact center operators, was used to organize the workstations of operators and supervisors. "Дом.ru"

The main objective of CTI in the implementation of the project is to replace the existing individual customer call centers with a comprehensive solution with centralized management, unified statistics and load balancing between remote operator sites.

Project Outcome

The solution introduced by CTI on the Cisco platform optimally distributed the load of operators, reducing the average waiting time of the subscriber. Through the project, Дом.ru significantly improved the quality of service and increased customer loyalty.

"Based on Web 2.0 technologies, the Cisco Finesse solution is the latest version of the contact center operator workstation. Cisco relatively recently began to offer its Customers to use this variant of Agent Desktop, and the installation in the project for "Дом.ru" has become one of the largest in the world, "said Pavel Teplov, Cisco Business Development Manager

Anatoly Guryev, director of the joint contact center "Дом.ru," emphasized: "Now there are first results from the transfer of contact centers to a new platform. We have achieved stable and high-performance operation of the hardware and software complex, extensive capabilities to control the quality of call processing and statistical processing of its results. "

"For CTI, the implementation of this project has become a very interesting task. Several systems have been integrated into a single infrastructure aimed at the continuous delivery of high-quality services. The project team of CTI specialists once again proved that our company is an honored leader among Cisco partners in the implementation of solutions for contact centers, "said Platon Begun, head of the Contact Centers department of CTI