AzureFeeds

We just announced NLWeb at Microsoft Build—starting with a GitHub repo to help developers explore it, and a short list of enterprise pioneers testing it out in the real world. Eventbrite is one of those early innovators shaping where this goes.

Known for their global events platform that helps millions discover and attend unique experiences, Eventbrite is now exploring how NLWeb can make that discovery process even more personal, conversational, and precise. Whether you’re looking for a date night idea or planning your next creative outing, NLWeb enables natural, expressive search queries that understand intent—not just keywords.

Below, Eventbrite shares their journey piloting NLWeb: what made them say yes, how it’s performing in early tests, and what they see on the horizon.

Q1: What inspired Eventbrite to try NLWeb?

When discovering things to do on Eventbrite, consumers are often looking for a range of activities that match the vibe of their group and occasion. Eventbrite offers an unmatched volume of unique, local experiences for every mood and interest. However, traditional search has required people to conform to a keyword-and-filter paradigm. Instead of offering rich context, many enter a single activity–or nothing at all.  

From AI-driven conversational experiences and research, we know that intent is centered around the “why” and “who” behind the gathering. People often have a range of potential activities and desired moods in mind. Their behavior is exploratory, their occasions are time-sensitive, and their preferences are flexible– something that shouldn’t be restricted by a search box.  

On the flip side, traditional search ranking methods aren’t well-equipped to understand the meaning behind more natural language expressions. We believe an NLWeb solution can better inspire, interpret, and effortlessly surface the most relevant options with remarkable speed, transforming how today’s consumers discover exactly what they want.

Q2: How did the setup process go for your team?

The setup process was smooth overall. Our team has extensive experience working with LLMs, so the primary focus was on applying our domain-specific knowledge and tuning the prompts and data inputs accordingly. NLWeb significantly accelerated our ability to launch customer-facing experiments by leveraging the deep language understanding inherent in LLMs. 

We were pleasantly surprised by the out-of-the-box query performance–both in terms of relevance and latency–which exceeded our initial expectations. As we continue to iterate and refine our approach, we anticipate even greater performance gains that could unlock new use cases previously considered out of reach for LLM-based solutions. 

Q3: What’s a query or interaction that made NLWeb really click for you?

Traditional site search is great at ranking specific activities like “paint and sip” or “bar crawl.” It also leverages proprietary event and customer data to provide recommendations for general queries like “fun events this weekend.” NLWeb excels at interpreting more nuanced, conversational queries that require deeper semantic understanding–such as “what should I do on a first date on a rainy day that doesn’t involve drinking” or “I want to find something to do that’s chill, very low key, to get to know new people–maybe be creative.”  

These interactions are particularly compelling because NLWeb doesn’t just match events to queries; it explains why these results align with what someone is looking for. By providing context and rationale behind each recommendation, NLWeb creates a deeper, more meaningful connection with people. This transparency builds trust, reinforcing our position as the essential, go-to resource for discovering memorable IRL experiences tailored to individual needs and preferences. 

Q4: How are you blending NLWeb with your current design or customer journey?

We’re eager to see how this new experience performs compared to traditional search. To start, we’re testing NLWeb as an alternative path to search to evaluate which types of queries it handles best and where it may need further tuning. We’re prioritizing experiments with new modalities using this technology and taking a customer-centered approach to learn what resonates most for them. 

Q5: If NLWeb reaches its full potential, what could that unlock for Eventbrite users or the web at large? 

We’re on the cusp of transforming how people search and discover. Search engines have trained people to interact with the web in a limited way. Traditional search and filters are restrictive tools for expressing intent and narrowing down results. With NLWeb, people can more freely express their desired outcome and let the product adapt accordingly.  

The user experience will evolve away from tedious filtering and keyword refinement toward something more like an exchange with a knowledgeable friend–someone who, with a bit of context, offers trusted recommendations backed by clear rationale.