jeecgboot/jimureport

JimuReport – A Free, AI‑Enhanced Visual Analytics Platform

What it is – JimuReport (also called 积木报表) is an open‑source, web‑based visual‑analytics suite. It lets you design traditional tabular reports, dashboards, large‑screen “big‑screen” displays, and a conversational BI interface, all through a drag‑and‑drop, Excel‑like editor.

Key capabilities

• Report designer – unlimited rows/columns, grouping, cross‑tabs, master‑detail, multi‑sheet, QR/barcode, conditional formatting, printable layouts, and data entry forms.
• Dashboard / big‑screen builder – Word‑style canvas with 28+ ECharts visual types, grid‑based layout, map components, video/iframe, and responsive mobile support.
• JimuChatBI – a Chat2BI feature that accepts natural‑language prompts (or screenshots) and, using an LLM (Claude Code or any OpenAI‑compatible model), automatically creates or edits reports, dashboards, and big‑screens.
• AI‑assisted generation – one‑sentence prompts can produce a complete report, dashboard, or KPI board; the AI can also suggest schema changes and fill tables with sample data.
• Spring Boot integration – starter artifacts (jimureport‑spring‑boot3‑starter, jimubi‑spring‑boot3‑starter, jimuchatbi‑spring‑boot4‑starter, etc.) let you embed the engine in any Spring Boot application.
• Broad data‑source support – 30+ connectors, including MySQL, Oracle, PostgreSQL, SQL Server, Hive, ClickHouse, Elasticsearch, MongoDB, CSV/Excel files, and Chinese‑government “信创” databases (达梦, 人大金仓, 神通, 华为高斯, etc.).
• Deployment options – run from source, Docker, or a “green” no‑install package; requires JDK 17+ and a MySQL 5.7+ instance (Redis optional).

Why it matters for AI/ML practitioners

• The platform demonstrates how LLMs can be plugged into enterprise BI tools to turn natural language into structured visualizations, a concrete example of generative AI for data analysis.
• Its open‑source “Skills” repository shows how custom Claude Code plugins are built, offering a template for extending AI‑driven analytics.
• By supporting any OpenAI‑compatible endpoint (DeepSeek, OpenAI, Azure, etc.), developers can experiment with different model families without changing the product.

Getting started

1. Add the appropriate Maven starter to your Spring Boot project (choose the version matching your Spring Boot major version).
2. Configure the AI endpoint in application.yml (base‑url, API key, model, etc.).
3. Run the example app (jimureport‑example) or launch the Docker image to explore the web UI.
4. Use the built‑in AI button in the designer to type prompts like “Create a department‑wise salary summary report” or upload a screenshot to have the system reconstruct it.

License & commercial use – The core code is LGPL‑licensed; a commercial license is available for those who need to remove the JimuReport branding.

---

All details are taken directly from the repository’s README; no additional features are inferred.