[Edaily Reporter Lee So-Hyun ] Naver Maps is introducing an “Agent AI” feature that handles everything from location searches to reservations through conversations with artificial intelligence (AI). Users no longer need to repeatedly search for exact business names or keywords; instead, by entering their desired situation and conditions in everyday language, the AI recommends suitable locations and, through further conversation, connects them to the reservation process.

Naver (NAVER(035420)) announced on the 17th that it has launched “Place Agent,” a feature that allows users to search for and book locations interactively within the map app.

Place Agent AI is an AI-based conversational search feature specialized for location discovery. It analyzes the user’s conversational context and location based on information about restaurants, cafes, bakeries, hotels, vacation rentals, hair salons, veterinary clinics, and other venues registered on Naver Maps to recommend suitable locations.

With traditional map searches, users had to search for specific keywords—such as business names or menu items—in multiple separate searches and then compare the results themselves. Place Agent is designed to take multiple conditions entered in a single sentence and reflect them all together to present suitable options.

For example, if you enter, “A Korean restaurant in Yongin with ample parking that’s good to stop by on the way back from visiting a grave,” it will recommend a location by synthesizing your travel context, type of cuisine, area, and parking conditions.

It also understands everyday expressions such as “Recommend a bakery tour route in Seongsu-dong,” “A quiet café in Yeonnam-dong where I can read a book alone on a rainy day,” and “A great tofu restaurant in Ui-dong to visit after a fall foliage hike on Bukhansan Mountain.”

Recommendations include reasons based on reviews and blog posts

Place Agent’s responses incorporate not only basic place information held by Naver—such as addresses, business hours, menus, and phone numbers—but also visitor reviews, blog posts, and reservation availability.

Rather than simply listing places, it also provides the characteristics of each location and the reasons for the recommendation. This approach synthesizes relevant information into a single response to make it easier for users to review the recommendations and make a choice.

Users can also ask situation-based questions—such as “a safe bet for a date” or “a place with easy parking”—without specifying a particular area.

It also supports “multi-turn” conversations where users can further narrow down the recommendations. For example, after receiving recommendations for several places, if a user asks, “Which of these is quiet?” the system will narrow down the candidates again based on the previous conversation and results.

“Book

a table

for four at 12 p.m. tomorrow”… A seamless process that includes booking

If the desired location supports Naver Reservations, the conversation can flow seamlessly from search to reservation request.

When a user types, “Please make a reservation for four people at Restaurant XX for lunch tomorrow at 12:00 PM,” the agent summarizes the reservation details and seeks the user’s approval. If the user accepts, the process proceeds to the reservation request.

You can use Place Agent AI by tapping the Agent AI button that appears when you select the Naver Maps search bar. You can also access it through the “Ask the Agent AI” button, which appears when there are no suitable search results or recommended questions displayed in restaurant search results.

The service is available to logged-in users aged 14 and older, and a separate notice is displayed for answers generated by AI.

Choi Ji-hoon, Head of Naver Place Search, stated, “Place Agent is a service optimized for the map environment by combining rich Place information,” adding, “We will continue to enhance its features and usability so that users can discover their desired locations more easily and accurately, and naturally proceed from booking to an actual visit.”