AutoBI One‑Stop Data Visualization Platform: Architecture, Technical Highlights, and Use Cases

As enterprises grow, their IT environments become increasingly complex and data volumes explode, making rapid data integration, value extraction, and decision‑making critical. Traditional data‑board development at the company required a lengthy, multi‑step process involving data warehouses, front‑ and back‑end development, testing, and deployment, resulting in low efficiency and resource waste.

AutoBI is positioned as an easy‑to‑use, one‑stop visualization platform that enables business users to create dashboards through drag‑and‑drop, offers a rich component library with dynamic data interaction, automatically adapts to PC and mobile devices, and streamlines the entire data preparation, processing, and consumption workflow.

The overall architecture consists of three layers: a data layer supporting common data sources, a data‑service middle platform providing a unified model registration and service outlet, and a front‑end application layer with atomic UI components that can be extended and maintain consistency across devices.

Technical Highlights

3.1 Real‑time and offline data switching: a mapping model is built during data model registration to associate metrics and dimensions with both real‑time (e.g., Redis, ES) and offline sources, and the query processor resolves the appropriate source based on configured rules.

3.2 Data query processor: decouples the data‑interface layer from chart rendering, converts both real‑time and offline data into a unified two‑dimensional table format, defines a standard graphic rendering component specification, and uses an adaptable adapter to transform the data for libraries such as ECharts or Highcharts.

3.3 Integration capabilities: email subscription for dashboard distribution, server‑side screenshot service for long‑page captures on PC and mobile, and mobile integration via apps (e.g., Car‑People App) to present reports and dashboards.

Scenario Case – AutoBI Features

Supports nine chart types (line, bar, pie, funnel, bubble, etc.), customizable page layouts, extensive filter options (single‑select, cascade, time, constants), and multiple data sources including Redis, MySQL, SQL Server, Elasticsearch, and Kylin.

The dashboard editor is divided into three zones: left side for component selection, middle for chart placement, and right side for configuration (data model view, chart settings, axis, series, and functional options such as component linking and download).

Deployment and Outlook

AutoBI has been widely deployed across the company’s big‑data center, enabling rapid dashboard creation (1‑3 hours versus weeks of development) and improving user experience. Future plans include multi‑dimensional analysis, heterogeneous data‑source integration, statistical reporting, and richer chart types to further close the gap with industry‑leading products.