Low-Code Platform for Efficient Data Synchronization at iQIYI: Design, Challenges, and Solutions

To address the software crisis and improve productivity, low-code/no-code concepts have gained attention, with Gartner predicting 65% of applications will use low-code platforms by 2024; iQIYI leverages this trend to accelerate business data synchronization.

The typical data flow for a mobile app includes data production back‑ends, synchronization services, databases & caches, backend APIs, and the app front‑end, each with distinct responsibilities and performance considerations.

Key challenges encountered were manpower bottlenecks due to the growing number of sync services, tight iteration cycles that jeopardize quality, rising middleware maintenance costs, and extensive duplicate coding across projects.

Existing data‑sync products were evaluated but none fully met iQIYI's heterogeneous, extensible requirements; insights from DataX's plugin mechanism and Spring Cloud Data Flow's orchestration inspired the search for a low‑code solution.

The Quqiao platform was built with two core components: a visual management console for composing reusable logic nodes, and a sync engine that executes the assembled workflow. This design enables rapid, low‑code delivery while maintaining high availability.

Platform capabilities include a drag‑and‑drop editor, entity‑mapping templates for JSON‑to‑business data conversion, plugin management for custom logic nodes, middleware resource management with dependency tracking, and seamless integration with deployment, logging, and monitoring systems.

The sync engine packages required plugins, loads configurations from a central store, initializes resources, processes data through the node chain, reports heartbeats, and feeds metrics to monitoring services, ensuring reliable operation.

Since launch, Quqiao has supported over 20 sync services and 30+ application instances, dramatically shortening development cycles from hours or days to minutes; future work will focus on automatic code generation for data‑access layers, private‑cloud deployment, and building an internal data marketplace to further reduce duplication.