AI Eng

“At first, we just wanted to see how many customers ChatGPT could bring in.” That’s how agentMention began, a Ukrainian startup that helps companies track how often ChatGPT, Gemini, Claude, and other AI systems mention and recommend their brands. The idea arose even before AI search became a separate direction for marketers. The founders wanted to understand whether AI assistants could become a new channel for customer acquisition and how businesses could adapt to changing user habits. Over time, the GEO market began to form around this — optimizing the presence of brands in generative search engine responses. Today, the team develops the platform on its own, without external investment, and has 20 paying customers. The service analyzes which queries the AI ​​mentions companies for, compares them with competitors, tracks the sources of recommendations, and suggests ways to improve brand visibility. dev.ua spoke with the founder of agentMention, who wishes not to reveal his name due to corporate nuances from his main place of work, about how AI search analytics works, how much it costs to develop such a product, how a startup earns on subscriptions, and whether it is possible to build a business in a market whose rules are still being formed.

A few years ago, a significant part of a creative’s value was tied to what they could do with their hands: write a text, draw a visual, find the right reference, put together a storyboard, edit a video, build a complex scene in 3D. Today, some of this work can be done in hours, and sometimes minutes. And it would be strange to pretend that this will not affect the profession in any way. It will. It already does.

In the 3rd episode of our partnership project with Google Ukraine — "AI to the rescue" — we took a look "under the hood" of Daylight, a premium supplier of interior textiles and wallpaper. The company's CEO, Yulia Bakun, shared her practical experience of integrating Gemini into workflows and talked about how automation is changing the routine of the B2B segment.

After a few days at an AI conference, I found myself thinking that we might be asking the wrong question about AI and the future of PR. We keep asking: Will AI replace communications professionals? But what is much more interesting is this: why do we need PR in a world where AI can already do a significant part of the work we used to call PR?