Back to Blog

Building App Factory: Scaling React Native Apps with AI & Google Play Automation

•Emmanouil Athanasopoulos•3 min read•595 words
React NativeNode.jsGoogle Gemini AIGoogle Play APIAutomationWebSockets

The Motivation Behind the Project

Managing and scaling dozens of React Native applications on the Google Play Store quickly becomes an administrative nightmare. From generating localized store descriptions across 49 markets to auditing release AAB bundles, verifying keystores, and tracking live subscription billing and AdMob revenue, manual maintenance drains valuable engineering time. App Factory was built to solve this exact problem: a unified command center that automates the entire React Native lifecycle—from AI spec generation to live Google Play Console telemetry sync.

Core Features and Design Goals

  • Automated AI ASO & Translation Engine: Leverages Google Gemini AI to generate high-conversion store copy and stream transcreated listing batches across 49 global Play Store markets in real time.
  • Local Codebase & Keystore Auditing: Scans local React Native project directories, verifies Android package IDs, audits release keystores (*.jks/*.keystore), and validates compiled production release AAB binaries.
  • Google Play Console Direct Billing (API v3): Operates without third-party billing wrappers. Parses auto-renewing subscriptions, catalog prices, and duration schedules directly via `androidpublisher.monetization.subscriptions`.
  • Cloud Storage Sales Report Sync: Automatically parses Google Play's automated GCS CSV sales report buckets (`pubsite_prod_rev_...`) for daily and monthly dollar earnings telemetry.
  • Real-Time WebSocket Streaming: Built on Node.js and Express with a bidirectional WebSocket layer (`ws://localhost:3001/ws`) to stream step execution progress and logs to a React 19 glassmorphism dashboard.

Deep Dive: How It Works Under the Hood

One of the biggest technical challenges in building App Factory was handling Play Store billing and monetization telemetry without relying on third-party SDKs like RevenueCat. Standard Google Play Developer API v3 endpoints for listing in-app products were deprecated by Google (returning 403 errors). To overcome this, I architected a dual-layer approach: (1) Using `publisher.monetization.subscriptions.list` for direct SKU catalog verification and duration parsing (`/ mo`, `/ yr`), and (2) Integrating Google Cloud Storage (`@google-cloud/storage`) to stream automated CSV sales reports directly from the developer's Play Console financial bucket. On the frontend, a React 19 dashboard receives real-time updates via WebSockets whenever telemetry or pipeline jobs run.

Technical Implementation

  • Built the backend engine using Node.js, Express, and native WebSockets (`ws`) for low-latency job dispatching and live status streaming.
  • Integrated Google Gemini API (`gemini-flash-latest`) for multi-locale ASO transcreation streaming across 49 Play Store regions.
  • Utilized Google Play Developer API v3 (`googleapis`) for direct customer review synchronization and store listing edits.
  • Engineered persistent local storage in `data/apps.json` with fail-safe DB rehydration to preserve verified sales metrics across server restarts.
  • Designed a responsive React 19 + Vite dashboard featuring dark-mode glassmorphism, per-app analytics tables, and real-time execution visualizers.

Architecture Overview

App Factory consists of a Node.js Express backend service (Port 3001), a WebSocket server for real-time telemetry streaming, a local file-system scanner watching React Native codebases (`PROJECTS_ROOT`), and a React 19 glassmorphism frontend (Port 5173).

Challenges I Faced Along the Way

A primary challenge was maintaining zero-mock data integrity. Play Store titles, customer reviews, ratings, and subscription plans needed to reflect real Google Cloud data. Bypassing deprecated API v3 endpoints while maintaining automatic catalog baseline calculations required extensive testing against live Google Cloud IAM service accounts.

The Technology Stack

Built with Node.js 20, Express, React 19, Vite, TypeScript, TailwindCSS, Google Gemini AI API, Google Play Developer API v3, and WebSockets.

What I Would Do Differently Next Time

Building internal developer tools and automation engines requires focusing heavily on resilience and zero-friction DX. Combining AI content generation with direct Google Cloud APIs transformed app management from hours of manual work into a single automated pipeline run.

Final Reflections

App Factory demonstrates how AI and direct API integrations can turn complex app portfolio management into an automated, single-pane-of-glass operation.