Tourism platform·2026

Visit Almaty (NTA) — Almaty's tourism platform

Almaty's tourism showcase, live in pilot mode at nta.visitalmaty.kz: places, events, mountain trails, bookings and city transport — on one map and in 6 languages. The catalog holds 2,900+ places, out-of-town trails with distance, elevation gain and time, city guides, bookings for stays, tables, tours and tickets with no prepayment, scooters, taxis and car rental, plus a city wallet with cashback and partner discounts. Businesses can list and claim their own pages. Under the hood it is a multi-tenant AI-first marketplace on Fastify + tRPC and Next.js 15 with pgvector semantic search.

Next.jsFastifytRPCPrismaPostgreSQLpgvectorClaudei18n
Visit Almaty (NTA) — Almaty's tourism platform

Project goals

  • Give visitors everything about the city in one place: places, events, trails, transport and bookings
  • Make the platform accessible to foreign guests — 6 languages, including Hindi, Chinese and Arabic (RTL)
  • Let local businesses list and claim their own pages
  • Build in multi-tenancy so new cities and brands can launch on the same platform

Tasks

  • Design a monorepo: a Fastify + tRPC v11 backend and a Next.js 15 frontend
  • Import and normalize a catalog of 2,900+ places with categories and a map
  • Build mountain trails with stops, distance, elevation gain and difficulty
  • Implement bookings for stays, tables, tours and tickets with confirmation and no prepayment
  • Integrate transport: scooters, taxis and car rental without a local phone number
  • Design a city wallet: coins, 5% cashback and partner discounts
  • Wire up PostgreSQL + pgvector for semantic search and Claude for translation and moderation
  • Set up deployment via Docker Compose behind a Caddy reverse proxy on visitalmaty.kz

Results

Measurable performance indicators

  • The platform is live in pilot mode at nta.visitalmaty.kz
  • A catalog of 2,900+ places: nature, city, culture, wellness, outdoor and transport
  • Mountain trails, city guides and bookings in a single interface
  • 6 interface languages, including an RTL layout for Arabic
  • A business dashboard: listing, page claiming and pricing plans
  • A multi-tenant core ready to launch the next cities

Key metrics

Specific performance indicators of the project

2 900+

Places in catalog

6

Languages

45+

Data models

pgvector

Semantic search