AI can already answer a question such as “Which laptop is best for remote work?” The harder part is making sure that answer reflects real products, current availability, exact configurations and the market the shopper is actually buying in.
That is the problem PriceHub.AI is trying to solve. The service combines large language models with the structured product infrastructure developed by E-Katalog over more than 20 years. Instead of asking an LLM to act as the database, the system uses AI as an interface to information that has already been normalised, matched and checked.
The underlying data infrastructure is already used by E-Katalog in Ukraine. Over time, the platform has built structured product catalogues, model-matching systems, retailer feeds and historical pricing data. That matters because online listings are rarely consistent: the same device can appear under different names, product codes or configurations across multiple shops.
PriceHub.AI adds a conversational layer on top of that foundation. A user can describe a real-world need rather than manually selecting technical filters. The system then has to translate the request into relevant requirements, identify suitable models and check current offers.
Technically, much of that work is coordinated by an AI orchestrator. Rather than sending every question directly to one language model, the orchestrator routes parts of the request to specialised modules. Some work with specifications and relationships between models, while others handle reviews, pricing and availability. The results are assembled before the language model generates a readable recommendation.
Localisation is another major part of the engineering challenge. E-Katalog has already adapted its catalogue infrastructure for the Polish market, but entering a new market involves more than translating an interface. Retailers, currencies, stock, product variants and pricing patterns differ from country to country, so recommendations need to be grounded in local market data.
This is where the company’s recent European Innovation Council funding becomes relevant. PriceHub Technologies was selected under the EIC Ukrainian Tech Call for its next-generation AI-driven product recommendation and price-comparison platform. The wider programme awarded €20 million to 41 Ukrainian technology companies, with individual grants ranging from €300,000 to €500,000.
According to project information published after the selection, PriceHub.AI received about €500,000. The funding is intended to support development of the technology core, local models and the orchestration layer, as well as prepare the platform for full market testing.
The UK is one of the important next steps. The wider ecosystem already operates E-Catalog in Britain, providing an existing local product and retailer database. PriceHub.AI plans to test the full service first in Ukraine, followed by Poland and the UK.
The same architecture is intended to support further expansion. E-Catalog also operates in the United States, where the volume of retailers, product variants and offers makes reliable model matching and data normalisation particularly important.
Another technical priority is explainability. PriceHub.AI is designed not only to recommend one or two products, but also to show why they fit the request, what trade-offs were considered and which information influenced the answer. Historical pricing can also add context when users want to know whether buying now makes sense.
The standalone PriceHub.AI service is still described by the company as an experimental prototype. The grant is therefore less about scaling a finished chatbot and more about strengthening and validating the infrastructure behind it.
That distinction may prove important. The next generation of AI shopping tools will not be judged only by how naturally they talk, but by whether their answers are grounded in identifiable products, current market data and transparent reasoning.
