Sometimes you find a product worth considering during a conversation. You look at a few options, ask more questions, compare details and gradually narrow things down.

On a traditional product comparison site like E-Katalog, the process usually starts with filters. You narrow down the models, explore their specifications, read reviews, compare prices and save interesting options to a list.

That approach still works well. But choosing the right filters isn't always easy, especially when you know what you need but aren't familiar with the technical details.

With PriceHub.AI, we're exploring a different starting point: a conversation supported by product data from E-Katalog and other sources. You can describe what you're looking for, ask follow-up questions and work through the options without having to figure out every specification in advance.

And as interesting products come up, why not save them straight from the chat? Not just to keep them nearby during the conversation, but to build a list you can return to later.

A small test with coffee machines

In our prototype, someone is looking for a coffee machine for cappuccinos and lattes, with easy cleaning as a priority.

The chat suggests two models: Philips 5500 LatteGo and De'Longhi Magnifica Evo. Both get saved to a list called Coffee machines.

A little later, another detail comes up: the person sometimes switches between regular and decaf beans.

This brings a third model into the conversation: De'Longhi Rivelia, with its interchangeable bean hoppers. We save it to the same list.

Here's how it works in our prototype.

One list, beyond one conversation

By this point, the list has three coffee machines that came up at different moments in the chat. It opens alongside the conversation, so you can select a model and see its photo, price and specifications without navigating to a separate product page.

We're also thinking about what happens after the conversation ends. A list might stay useful for weeks while you're deciding on a coffee machine, collecting gift ideas or choosing parts for a new PC. The goal is to make it easy to pick up where you left off, not just to keep products visible while you're chatting.

Traditional product pages, filters and comparison tables still have their place. We're exploring how they can work with the conversational experience, and there are open questions about how much detail belongs in the side panel and what someone should see when they return later.

For now, it's one small part of the shopping experience we're experimenting with.