Skip to content
PD
EN RU
Web products & EdTech Client: Owner of a rideshare service

A BlaBlaCar-Style Rideshare Service

A BlaBlaCar-style rideshare service with no install and data in Russia, in 6 weeks

A web app for finding travel companions in Russia, built for a client on the BlaBlaCar model. It opens from a link in the phone's browser, with no App Store, Google Play or install.

A trip with stops, the home screen with search, and the route on Yandex Maps
Built in
6 weeks
from the mobile template to launch, August - October 2026

Before

The client needed a rideshare service for Russia: it opens from a link on a phone, user data stays on a server in Russia under law 152-FZ, maps come from Yandex. All there was: an English mobile template on Flutter and Firebase.

What I built

The template was moved to the web and to Supabase on the client's server: booking a leg of a route, a per-leg seat price, routes on Yandex Maps, SOS, chat, push and a nine-section admin panel.

After

The service runs on the client's server over HTTPS, 6 weeks from template to launch. 390 automated tests cover the code, the main flows also run in a real browser, and the backup is restore-tested every week.

What the client got

  • Book a leg of the route

    Riders can get on and off at a stop along the way, and free seats are counted for every leg of the trip.

  • Per-leg pricing

    The seat price comes from distance and the car's rate, and the driver can adjust it. The rate is set in the panel.

  • Yandex Maps

    Address suggestions, a road route through every stop, and trips on a map with the price at each pickup point.

  • Safety

    An SOS button with coordinates, driver document checks and a "women only" filter that can be trusted.

  • Chat and push

    Driver-rider chat with read receipts, push on bookings and trip status, iPhone included.

  • Admin panel

    Nine sections, from users and complaints to rates and copy. SOS and complaint counters update on their own.

The home screen and a trip card where the rider picks where to get on and off
A road route through the stops, and trips on a map with the seat price
The app suggests towns on the way and prices a seat for every leg
The admin panel: every trip with its status, complaint and review counters in the menu
How it was built: details and stack

Tech Stack

Flutter WebDartGetXSupabasePostgreSQL 17 + PostGISYandex Maps JS APIWeb PushnginxClient's own server

The brief

The client needed a rideshare service for the Russian market. The conditions:

  • the app opens from a link on a phone, with nothing to install;
  • user data is stored on a server in Russia (law 152-FZ), so Firebase was out;
  • addresses, routes and the map come from Yandex Maps;
  • moderation and settings live in a separate admin panel.

The starting point was an English-language mobile template on Flutter and Firebase. It had to be moved to the web, onto its own backend, translated into Russian, and the trip logic had to be reworked.

What was built

Search and booking a leg

A rider searches for a trip by address, with suggestions. The results also include trips that pass through the right towns on the way.

Riders can get on and off at an intermediate stop. Free seats are counted per leg: a rider from Tver to Bologoye does not take a seat on the Moscow - Tver leg. The trip card shows how far to walk to the pickup point and from the drop-off point.

Route and map

The route is built on Yandex Maps along the roads, through every stop. A separate screen shows trips on the map, with the price at each pickup point.

Publishing and per-leg pricing

The seat price is calculated for each leg from the distance and the car’s rate, and the driver can adjust it. The app suggests towns on the way by itself, skipping garden plots and cottage villages. The rate per kilometre is set in the admin panel.

LegDistance, kmSeat price, ₽
Moscow - Tver185925
Tver - Bologoye191955
Bologoye - Saint Petersburg3611805

The rate in this example is 5 ₽ per km.

Admin panel

The panel has nine sections: users, trips, complaints and SOS, reviews, support, payouts, cars and prices, texts, and settings. The counters for new SOS signals, complaints and documents awaiting review update without a page reload.

Safety, notifications, sign-in

  • The SOS button sends the phone’s coordinates during a trip. If location access is denied, an approximate point from the network is used. The signal reaches the panel with the participants’ phone numbers and a point on the map.
  • Driver document checks. Once checked, the gender in the profile cannot change, so the “women only” filter can be trusted.
  • Riders’ reviews of a trip and drivers’ reviews of a rider are kept apart in the profile and in the admin panel.
  • Driver-rider chat with read receipts.
  • Push notifications about bookings, messages and trip status, iPhone included (the site is added to the Home Screen).
  • Sign-in with a password or a link from an e-mail. The e-mails are in Russian and go out through the client’s own mail server.

How it works

Firebase was replaced with Supabase: PostgreSQL with PostGIS to search trips by distance, row-level access rules, Realtime for chat, storage for photos, and server functions for push and sign-in.

Everything runs on the client’s server without Docker and without root access. The database and services were built from source and official builds, and a watchdog restarts any service that goes down.

A backup is made every night. Once a week the latest backup is restored into a temporary database automatically, and a check confirms every row is there.

Trips that never started drop out of search on their own after a day.

PartTechnology
AppFlutter Web, Dart, GetX
ServerSupabase: PostgreSQL 17, PostGIS, GoTrue, PostgREST, Realtime, Storage, Edge Functions; nginx
MapsYandex Maps JS API
Notifications and mailWeb Push, SMTP
Admin panelPlain JavaScript, no framework

Result

The service runs on the client’s server over HTTPS. Going from a mobile template to a finished web service took 6 weeks. 390 automated tests cover the code, and the main flows also run in a real browser.

The screenshots use demo data.

Services used in this project

Related case studies

Need a similar system?

Tell me about your challenge - I will propose an architecture and the shortest path to a working product.