AI is changing the world, and those who adapt to the changes will benefit

Artificial intelligence is no longer just a technology of the future – it is a reality that changes business, everyday life and even creativity. But how fast is the AI ​​project market growing? Which segments are developing the fastest and what challenges do companies face when implementing AI solutions, said Alexey Pilipchuk, Technical Director of Softline Solutions.

The AI ​​project market is growing rapidly. Which segments are developing the fastest?

The AI ​​market is truly exploding. We see new solutions emerging in both public and local segments. AI is already being used to create films, automate routine tasks, and even in everyday life – from smart assistants to data analysis systems. If we talk about segments, almost all areas demonstrate growth. However, global experience shows that the market for AI equipment stands out in particular.

What global factors have the greatest impact on the AI ​​market?

The main factor is data. The world has accumulated colossal amounts of information that require processing. Increased computing power and growing demand for automation, including routine operations, create ideal conditions for the development of AI. This gives a powerful impetus to its implementation in a variety of areas - from medicine to retail.

How have customer needs changed over the past 2-3 years? What requests have become dominant?

Customers have become more pragmatic. They want to see breakthrough technologies, but at the same time they require proven cases. The main question I hear is: “Where does this already work?” – Not everyone is ready to be a pioneer.

Today, AI is seen as a tool for optimizing routine processes. For example, call centers are actively introducing “robots,” and editors are being replaced by platforms that analyze news from open sources and automatically generate texts.

What are the main challenges customers face when implementing AI solutions?

The main problem is the underestimation of AI potential. Customers are often not ready to spend time on fully mastering the technology. In addition, the market lacks specialists who can not only implement solutions, but also integrate them with existing systems.

Another big issue is data security. Many are afraid of leaks, but are not ready to invest in the infrastructure needed to run local models.

What AI technologies are most in demand today? Why?

Computer vision is in first place. It is used to recognize images and videos, for example, in anti-drone protection systems or to analyze documents in retail.

The second direction is natural language processing. These are chatbots, voice assistants, and tone analysis systems. Almost every major bank already uses AI assistants to answer typical questions.

In general, AI actively solves problems related to routine operations, and this is its main value.

How do vendors adapt their solutions to customer needs?

Almost every vendor is implementing AI in their products, although so far more often in the role of an assistant. For example, one vendor added a tool for automatically compiling meeting minutes to its videoconferencing solution.

However, deep adaptation is still rare. More often, standard "boxed" functions are offered, which do not always meet the specific needs of customers.

Read more materials on this topic in Compass CIO

Is there a gap between expectations and reality?

Yes, and it is quite significant. Many customers dream of a “magic button” that will solve all problems. This leads to an overestimation of the capabilities of AI and an underestimation of its limitations.

For example, in one project, the customer wanted a computer vision system to perform complex analytics with minimal computing power. This is physically impossible, but such expectations are common.

What role do integrators play in the successful implementation of AI projects? What competencies are most important?

The integrator is a key player. He helps the customer understand how the AI ​​solution should work and what can be improved.

Key competencies of the integrator: deep understanding of AI technologies, experience in project management and skills in working with big data.

What are some examples of successful integration of AI solutions into complex systems?

One of the striking examples is chatbots that analyze the tone of communication, take into account the history of interactions in the CRM and offer personalized recommendations.

Another example is AI in information security, which identifies abnormal user behavior on the network and helps prevent threats.

What are the most pressing ethical and regulatory issues related to AI?

The main issue is the processing of confidential data. The development of local models and control methods partially removes this problem, but it remains relevant.

The question of responsibility is also important: who is responsible for the decisions made by the AI? For now, it is the human who must make the final decision, and the AI ​​acts as an advisor.

How to overcome technical challenges such as integration with legacy systems?

First of all, we need to recognize the need to transition to new technologies. A complete rejection of outdated systems is impossible immediately, but their gradual modernization taking into account compatibility with AI solutions is a realistic path.

There are also middleware solutions that allow you to integrate new technologies with legacy systems without completely replacing them.

What trends in AI projects will dominate in the next 3-5 years?

First, there is the rise of generative AI. Models will become better at creating text, images, videos, and even entire films.

Secondly, AI will become more accessible to the mass user. Solutions will appear that even small companies can use.

And of course, AI will become an everyday tool, its influence will only increase.

What advice would you give to customers and integrators for the successful implementation of AI projects?

First, you need to clearly define what the final result of the project should be and what data is needed for this.

The next step is to find a team with the necessary competencies. Talents can be both experienced specialists and young talents.

And finally, pay attention to the issues of integration and information security. Without this, the AI ​​solution will remain just an experiment, without any real benefit.