Ticksy Booking App

Ticksy Booking Application

Ticksy is a full-stack movie ticketing platform that combines a modern Next.js front end, a Spring Boot API, and supporting AI and data-ingestion services. The project demonstrates how a cloud-ready ticket booking experience can be assembled with modular services, event discovery, secure payments, and operational tooling.

Highlights

• Cross-service architecture that separates the UI, core booking API, generative AI assistants, and ingestion pipelines.
• PostgreSQL-backed catalogue with movie metadata that can be enriched and seeded via gRPC-based tooling.
• Integrated payments using Razorpay test credentials for a realistic checkout flow.
• Production-grade frontend built with shadcn/ui, Clerk authentication, TRPC, and TanStack Query for a polished user experience.

Repository Structure

Technology Stack

• Frontend: Next.js 15, React 19, Tailwind CSS, shadcn/ui, TRPC, TanStack Query, Clerk authentication.
• Backend: Spring Boot 3.5, Java 21, Maven, PostgreSQL, Razorpay SDK.
• AI & Tooling: FastAPI, LangChain (OpenAI, Google, Groq connectors), Gradio, Uvicorn, Pinecone, Sentence Transformers, gRPC.
• Infrastructure: Docker Compose for PostgreSQL, uv/virtualenv for Python environments.

Prerequisites

• Docker & Docker Compose
• Node.js 18+ and npm (or pnpm/yarn) for the frontend
• Java 21 and Maven 3.9+ for the Spring Boot API
• Python 3.12+ (AI service) and Python 3.13+ (data-ingestion) – uv is recommended for managing isolated environments
• Razorpay test keys, Clerk keys, and any third-party API keys required by .env files for the frontend/backend/AI service

Getting Started

1. Clone the repository

   git clone <repo-url>
   cd ticksy-booking-app
   

2. Start PostgreSQL

   docker compose up -d
   

   The database is available at postgres://postgres:postgres@localhost:5432/mydb.

3. Seed movie data (optional but recommended) Use the helper commands in command.md to truncate or seed tables. If you maintain a movies_seed.sql file (or a similar dataset), pipe it into the running container:

   cat movies_seed.sql | docker exec -i ticksy-booking-app-postgres-1 psql -U postgres -d mydb
   

Run the Backend API

cd book-my-show-be
./mvnw spring-boot:run

The API will expose REST endpoints (and Web APIs consumed by TRPC) on http://localhost:8080.

Run the Frontend

cd book-my-show-fe
npm install
npm run dev

The Next.js app defaults to http://localhost:3000. Ensure the .env.local file contains the Clerk publishable key, backend base URL, and Razorpay key ID.

Run the AI Assistant Service

cd book-my-show-ai
uv sync  # or: pip install -r requirements.txt
uv run uvicorn src.main:app --reload

The FastAPI service powers conversational helpers and can be mounted behind the frontend or tested directly at http://localhost:8000/docs.

Run the Data Ingestion Pipeline

cd data-ingestion
uv sync
uv run python main.py

This project handles ingesting external catalogues, compiling protobuf contracts (compile_proto.py), and pushing embeddings to Pinecone.

Database Tips

Useful psql helpers from command.md:

docker exec -it postgres_db bash
psql "postgresql://postgres:postgres@localhost:5432/mydb" -f movies_seed.sql  # replace with your seed file
# Truncate bookings and reset identity
docker exec -i ticksy-booking-app-postgres-1 \
  psql -U postgres -d mydb \
  -c 'TRUNCATE TABLE "movies" RESTART IDENTITY CASCADE;'

Adjust container names if your Docker Compose project name differs.

Payments

Checkout uses Razorpay test mode. Use the following card to simulate successful payments:

Troubleshooting

• Database migrations failing: Ensure Docker containers are running and that the application can reach localhost:5432.
• TRPC/Clerk errors: Double-check environment variables in the frontend .env.local file and that Clerk is configured for the correct domain.
• Python dependency conflicts: Use uv or separate virtual environments for book-my-show-ai and data-ingestion; they target different Python versions.

Contributing

1. Fork the repository and create a feature branch.
2. Follow the linting or formatting tools configured by each service (for example npm run lint for the frontend or Maven/ruff tasks if added).
3. Submit a pull request detailing your changes and include screenshots when UI updates are involved.

Enjoy exploring the Ticksy Booking Application and feel free to extend it with new ticketing, personalization, or analytics features!