Wegosty – A Gastro-Navigation App with Native Booking and an AI Concierge

After launching the B2B platform for venues, Wegosty expanded into a full gastro-navigator. The goal was clear: bring together the best gastro-tourism activities in one app — from restaurants and cafes to wineries, farms, and agro-estates. So that a user doesn't just find a place, but immediately books a table or tour without leaving the app. Venue selection is deliberate — listings can't be bought, Wegosty has its own methodology for choosing what makes the cut.

The app is built for urban audiences and tourists:

• young families looking for cafes and kids' activities
• couples planning a dinner or a wine tasting
• gastro-tourists exploring regional cuisine
• spontaneous visitors deciding where to go right now

Unlike the B2B platform, the focus here is on the end guest. The interface is deliberately visual: a map, bold venue cards, curated collections, and quick access to booking.

Development ran in parallel with the Wegosty platform to keep both products compatible. We started by mapping out the core scenario: a user should see venues around them, filter by category (restaurants, cafes, wineries, farms, museums, tours), browse cards, and go straight to booking through the embedded form.

Collections played a central role — a thematic gastro-guide for the city and region:"where to go with kids", "wine routes", "top local dishes","new openings this week". They let users quickly find ideas and plan their day.

The app includes:

• a map of venues and events with a wide range of customizable filters
• thematic collections and routes
• detailed venue cards with online booking, chef recommendations, and menus
• booking history and e-tickets
• saved places, grouped by city
• push notifications about upcoming visits and promotions
• integration with the Wegosty booking form and AI receptionist

The user starts on the map, where categories are displayed: restaurants, cafes, bars, wineries, farms, kids' activities, museums, tours, masterclasses. Each venue type has its own pin color, so you can orient yourself without extra filtering.

Polina is visiting a new city for the weekend. She opens Wegosty and sees the "Gastro Weekend" collection. She picks the nearest agro-farm, opens the card, checks the tasting schedule, and taps "Book" — the app opens the Wegosty booking form.

If Polina has questions, she can message the venue — the AI concierge responds, suggests available slots, and helps complete the booking. She gets an e-ticket with a QR code and a push notification an hour before the tasting.

On the map she finds a cafe on the way from the winery, saves it to favorites, and builds a route. That evening she books a table at a local cuisine restaurant through the same app and orders dishes from the chef's recommendations she'd seen in the venue card.

The app's design is built around three scenarios:

• find a nearby venue quickly
• pick a relevant collection or event
• book a table or ticket in a couple of taps

The interface is visual: venue photos on the map, venue cards, categories, large images of dishes and interiors. We paid attention to navigation: the map isn't cluttered, filters are simple, and cards contain only the information needed to make a decision.

Built-in analytics tracks key actions: venue card views, collection opens, transitions to booking, app installs, and user retention.

This data helps the team develop the product: understanding which venue categories are most popular, which collections perform best, where users convert to booking and where they drop off. Event-based analytics lets the team see audience movement in real time and adjust the interface accordingly.

React Native — cross-platform mobile development

Firebase — push notifications

WebView — embedded booking via the unified Wegosty form

Backend — Laravel + PostgreSQL

Mapbox — maps

Imgproxy — image optimization

The main engineering challenge was connecting the mobile app to the B2B platform cleanly across different budget phases. In the first iteration, booking is handled through WebView — this saved resources without reducing functionality for the user. The architecture is already set up for transitioning to a native in-app booking flow.

A separate task was building the map with color-coded venue types and smart filtering: selecting a category removes all other pins from the map without cluttering the interface. For city-grouped saved places, we built a local storage system that lets users return to their saved venues without being logged in.

The app was designed as bilingual from the start — Russian and English, detected automatically from the device language. This matters for gastro-tourism regions that receive international visitors.

Next steps include developing recommendation algorithms, personalized suggestions for dishes and drinks (not just venues), improved routes, and deeper work on the AI concierge. A venue news feed and a gastro-tourism pass are also planned.