Andrej Judiak


Andrej Judiak
Andrej Judiak thinks about on-site search the way a shopper does, not just a vendor.
As Head of Benelux at Luigi’s Box, he helps online retailers and B2B sellers turn better search and product discovery into real business results.
He specializes in AI-powered search and product discovery, and in how recommendations should differ by vertical: size and gender filters for fashion, compatibility over similarity for electronics, easy reordering for groceries, price segmentation for jewelry.
How product discovery is changing thanks to AI
Search used to mean typing the exact product name and hoping for the best. Now shoppers ask for “something warm for hiking” or “a guitar like the one George Harrison played,” and expect the right products to show up anyway.
In this session, Andrej Judiak looks at how AI is reshaping product discovery: search that understands typos, vague descriptions, and full sentences; recommendations that adapt to what a shopper is doing right now, not just their purchase history; and merchandising that no longer means dragging products around a category page by hand. Expect real, on-site examples, not just theory.
Five questions for Andrej
1. What is one development e-commerce companies can no longer ignore?
Search and product discovery that understands full sentences, vague queries, and prompts from AI agents – not just keywords – is no longer optional. Shoppers today type the way they would ask a friend or an AI assistant, and if a store’s search only matches exact keywords, it loses sales before a visitor even reaches a product page. This has stopped being a back-office technical detail and has become a core USP, something that decides whether shoppers find what they want, trust the store, and convert. The retailers who get this right now will have a real head start, because better search means more usable data, more data means better recommendations, and better recommendations mean higher conversion rates and larger order values down the line.
2. What do you see as the next major shift in e-commerce?
Conversational search and agentic commerce. Buyers are already discovering, comparing, and in some cases even purchasing products through AI assistants and autonomous agents acting on their behalf, rather than typing keywords into a search box themselves. That means the queries hitting a store increasingly arrive as full sentences, vague descriptions, or structured prompts generated by another system, and the store needs to understand and answer them just as well as it answers a traditional keyword search. Retailers who treat this as a future problem risk becoming invisible to an entire emerging channel of demand, while the ones who adapt early will capture buyers before their competitors even realize the channel exists. It is a genuine shift in how discovery happens, not just a new marketing buzzword, and it will separate the stores that grow from the ones that quietly lose relevance.
3. Can you give us a sneak peek of what visitors can expect from your session?
Expect a live, hands-on demo of KATO (knivesandtools.nl) / Luigi’s Box AI rather than slides full of theory. I will show how we understand natural-language, full-sentence, and agent-driven queries straight out of the box, with no months of prior data collection or manual relevance tuning needed before it starts performing.
We will look at real examples from live e-commerce sites, including vague and messy queries that would trip up a traditional keyword-based search, and see exactly how the system handles them in real time.
My goal is for attendees to walk away not just believing this is possible, but having seen with their own eyes what it looks like running on an actual production store.
4. Which customer expectation has changed the most over the past few years?
Shoppers now expect on-site search to understand what they actually mean, even when they type full sentences, vague descriptions, or half-remembered product names, because that is the experience they already get every day from Google and AI assistants like ChatGPT.
A few years ago, people were used to typing careful keywords and scrolling through irrelevant results; today, if a store’s search cannot keep up with a conversational query, shoppers simply assume the store does not have what they are looking for and leave, often without ever realizing the product was actually in stock. On top of that, they expect this level of understanding to work well from day one, not after months of the retailer tuning synonyms and relevance rules behind the scenes. That shift in patience and expectation is exactly why search and discovery has moved from a technical detail buried in the tech stack to a business-critical part of the customer experience, and why it now sits on the agenda of commercial leaders, not just developers.
5. What is your most important piece of advice for e-commerce professionals preparing for the future?
Compete on total return on investment, not on a feature checklist. When you line up serious vendors side by side, the feature differences are often surprisingly small, so chasing every capability a competitor lists is usually the wrong use of time and budget. The real advantage comes from speed to value and price-to-performance: how quickly a solution starts delivering results on your own site, and how much of that value you keep after accounting for cost and effort. My advice is to pressure-test any vendor or internal project against those two dimensions before anything else, and to be honest with yourself about how long you can really afford to wait for a project to pay off. In a market moving this fast, the teams that win are the ones that get real value in weeks, not the ones with the longest feature list on a slide.