Neural networks will be abandoned: companies will transfer customers to almost complete communication with AI in three years

In two or three years, Russian companies will transfer their customers to almost complete communication with artificial intelligence. More than 50 companies in Russia are already testing virtual assistants for customer and technical support, Izvestia has learned. These are not just assistants, but smart systems that learn by themselves to recognize the essence of requests and instantly provide ready-made answers and solutions. AI will solve problems in 95% of cases, and only the most difficult cases will remain for humans.

What tasks will be transferred to artificial intelligence

Over the next two to three years, most of the tasks of operators will be transferred to intelligent systems - more than 50 Russian companies are already testing virtual assistants for working with clients and technical support. We are talking not just about chatbots, but about advanced solutions that are able to independently analyze requests, find the essence in them, instantly generate responses and offer solutions based on internal documentation. Only the most difficult cases will remain for operators, Dmitry Chernous, head of technology consulting at MTS AI (part of MTS), told Izvestia.

According to him, such technologies will be able to fully automate typical tasks, reducing the processing time of requests by at least 20%, and AI assistants will perform from 50 to 70% of the work that support staff are currently doing.

— The first-line support roles will disappear — those who sign up clients for services, register requests and answer simple questions. For example, AI agents will be able to connect to users' calendars and external services themselves in order to automatically record clients," Dmitry Chernous said.

For non-standard or complex requests, AI operators will offer ready-made algorithms for actions: step-by-step instructions for returning goods, troubleshooting technical problems, or handling complaints. This will reduce the burden on operators and reduce the number of errors, the expert emphasized. AI operators will also be able to record all interactions between the user, the operator, and the system, which will allow them to evaluate work efficiency, identify typical problems, and improve the model based on real data, Dmitry Chernous noted.

— Despite the development of voice technology and emotional synthesis, personal interaction will still be necessary in difficult cases. Restoration of trust, apologies, empathy, support in crisis situations — all this will remain the sphere of a person, the so—called emotional support," he said.

Dmitry Markov, CEO of VisionLabs, believes that approximately 95% of communication between people and companies providing services will be carried out using artificial intelligence technologies. According to his forecasts, this will happen by the end of 2026.

The robots that exist now do not always respond clearly when answering user questions. Many requests go unanswered, which forces people to ask for a conversation with a human operator.

Can AI replace operators

The advantages of using AI assistants are obvious: This means round-the-clock availability and the ability to process thousands of requests simultaneously without loss of quality. Artificial intelligence does not get tired, does not require vacations and is not subject to the human factor — it responds almost instantly even with a sharp increase in requests, said Vladislav Belyaev, executive director and co-founder of the AutoFAQ AI platform.

The introduction of digital assistants makes it possible to reduce the load on contact center operators by three to six times, and the average speed of solving typical requests increases up to six times, he noted.

"The saving of working time is about 20-30 hours per month per employee, which is especially important for companies and services whose support workload amounts to tens or hundreds of thousands of conversations per month," the expert said.

He stressed that the use of AI assistants in support services is not a matter of replacing people, but of optimizing processes and working more quickly. Digital assistants open up new customer interaction opportunities for customer support services. And these prospects will lead to the development of service and increased efficiency. Modern digital assistants act as a partner for the operator.

Izvestia has sent inquiries to major Russian companies and banks — Yandex, Kaspersky Lab, T-Bank, Alfa-Bank, Sberbank, and others.

The VTB press service informed Izvestia that the work of qualified call center operators on non-standard, complex user issues will be organically complemented by the participation of digital assistants, who, thanks to the development of artificial intelligence technologies, will be able to solve an increasing number of routine client requests.

Daria Kasatkina, head of LLM automation at Avito's support service, told Izvestia that the company is testing the use of large language models in several areas of support: automating routine operations, improving the quality of communication with customers, and speeding up information search.

— Generative AI takes on mechanical tasks: data analysis, knowledge base search, query preprocessing, which allows operators to focus on solving complex problems. People began to process requests faster, they demonstrate high satisfaction with generative AI tools, which guarantees rapid technology adoption and return on investment," she said.

According to her, users receive faster and better answers, while maintaining a human approach to solving questions.

The transition to the widespread use of AI in customer support is a logical and, in fact, inevitable stage in the development of services, Anton Nemkin, a member of the State Duma Committee on Information Policy, Information Technologies and Communications, federal coordinator of the Digital Russia party project, told Izvestia. Chatbots do not always work correctly, users expect advanced solutions that can independently analyze requests.

Modern AI assistants have already gone beyond simple assistants. We are talking about systems that are trained on internal documents of companies, recordings of dialogues, and over time really "understand" the context and intentions of the user, he added.

— A complete replacement of a person is definitely impossible. AI copes well with repetitive tasks, but in situations where flexibility, emotional empathy or making non-standard decisions are required, live operators are still indispensable," the deputy said.

In addition, companies are required to ensure transparency: the client must understand whether he is communicating with a person or with a machine. It is necessary to provide in advance the possibility when the user may request to transfer the conversation to the operator. This is not only a matter of trust, but also an element of respect for the client. Proper division of responsibilities between a person and an algorithm is the key to quality service, he concluded.