Architecture

StorePush is built as a modern web application with a clear separation between the user interface, API layer, and background processing.

System Overview

User Request → API (tRPC + REST) → Background Job Queue → Apple API
                                         ↓
                                    PostgreSQL Database
                                         ↓
                                    Real-time Progress Updates

Core Components

Web Application

The frontend is a Next.js application with React Server Components. It communicates with the backend through tRPC for type-safe API calls.

API Layer

StorePush exposes two API interfaces:

tRPC — Used internally by the web application for type-safe, efficient communication
REST API — Auto-generated from tRPC routers via OpenAPI, available for external integrations

Both interfaces share the same authentication, validation, and business logic.

Background Job Queue

All Apple API operations run as background jobs in a PostgreSQL-backed queue. This design ensures:

Reliability — Jobs persist across server restarts
Ordering — Each app has its own queue, processing jobs sequentially
Observability — Progress is tracked in the database and visible in real-time

Apple API Integration

StorePush communicates with Apple through the App Store Connect API using JWT authentication. Tokens are generated on-demand and automatically refreshed before expiry.

Data Model

The core data relationships:

User (1) → App (N) → Version (N) → VersionLocalization (N)
                   → AppleApp (1)   (stores credentials)
                   → Build (N)
                   → Review (N)
                   → QueueJob (N)

Each User can have multiple Apps
Each App has one AppleApp record storing Apple credentials
Each App has multiple Versions, each with locale-specific VersionLocalizations
Builds, Reviews, and QueueJobs are associated with their parent App

Scheduled Operations

StorePush runs several cron-based operations:

Job	Schedule	Purpose
Pull Changes	Every 5/15/30 min (configurable)	Sync incremental changes from Apple
Pull Analytics	Daily at 02:00 UTC	Fetch analytics data
Predict Analytics	Sundays at 03:00 UTC	Generate AI performance predictions