DeepResearch

中文 | English

📖 Introduction

DeepResearch is an intelligent research Agent built on Spring AI Alibaba Graph, designed to tackle complex research tasks. It adopts a Multi-Agent collaborative pattern, supporting dynamic task planning and execution. The system integrates multi-source online search and Hybrid RAG technology, combined with Secure Sandbox for Python code execution, enabling efficient data analysis. Through Reflection, HITL, and Self-evolution Memory, the Agent can continuously self-optimize, ultimately outputting high-quality research reports with deep insights.

✨ Technical Features

• 📋Plan and Execute: Dynamic planning and automatic execution for complex problems
• 🤖Multi Agent: Multi-agent collaboration (e.g., Researcher, Coder)
• 🌐Online Search: Integrated multi-source search services including Tavily, Jina, Aliyun AI Search
• 📖Hybrid RAG: Combines vector and keyword retrieval for comprehensive information acquisition
• 🔄Reflection: Agent self-reflection for continuous output quality optimization
• 🚶‍♂️HITL: Human-in-the-loop feedback for enhanced controllability
• 🧬Self-evolution Memory: Self-evolving memory structure and user role memory based on interaction feedback
• 🖇️MCP Allocation: Support for MCP allocation in multi-agent scenarios
• 🔒Secure Sandbox: Secure Python code execution in Docker sandbox environment
• 📊Report Generation: Supports HTML report preview, Markdown, PDF and other report formats

🎋 Project Architecture

DeepResearch/
├──  ├── src/
│    ├── agents                          # Multi-Agent initialization, MCP allocation, observability initialization
│    ├── config                          # Graph construction, project Config classes
│    ├── controller                      # HTTP endpoints
│    ├── dispatcher                      # Graph EdgeAction
│    ├── model                           # Base project entities
│    ├── node                            # Graph key node definitions
│    ├── rag                             # RAG core implementation
│    ├── repository                      # Model configuration loading
│    ├── serializer                      # Message serialization implementation
│    ├── service                         # Business logic implementation
│    ├── tool                            # Agent Tool definitions
│    ├── util                            # Project utilities
│    └── DeepResearchApplication         # Application entry point
├──  ├── resource/                  
│    ├── prompts                         # Core prompts
│	 ├── mcp-config.json                 # Agent MCP configuration
│    ├── model-config.json               # Multi-Agent model configuration
├──  └── website-weight-config.json      # Search engine weight configuration

🧩 System Architecture

More illustrations

🔍 Running Example

Demo Video

🚀 Quick Start

Prerequisites

• Java 17+
• Maven 3.6+
• DashScope API Key

1. Clone and Build

git clone https://github.com/spring-ai-alibaba/deepresearch.git
cd deepresearch
mvn clean install -DskipTests

2. Configure API Key

export AI_DASHSCOPE_API_KEY=your-api-key-here

3. Start Application

Start from Project

Backend:

cd deepresearch
mvn spring-boot:run

Frontend:

cd ui-vue3
pnpm install
npm run dev

Docker Startup

• Build the Docker image from the project directory. This may take ~5 minutes depending on network speed.

cd deepresearch
docker build -t alibaba-deepresearch:v1.0 .

• After building, run the container and set environment variables:

docker run -d \
  --name alibaba-deepresearch \
  -e AI_DASHSCOPE_API_KEY="your_key_here" \
  -e TAVILY_API_KEY="your_key_here" \
#  -e JINA_API_KEY="your_key_here" \ optional
  -p 8080:8080 \
  alibaba-deepresearch:v1.0

• Alternatively, use docker-compose to start Redis, ElasticSearch, and the app:

  docker-compose up

💡Note:

• Set API keys in the .env file
• Config files are under dockerConfig; you can also set keys and related configs there

4. Configuration

• API KEY
• Search Services
• Storage Options
• Coding Node
• RAG
• MCP Service (WIP)
• Short Term Memory

5. Debug and Observability

Supports integration with Langfuse observability system. See documentation for configuration details.

Related API Documentation

• DashScope (Alibaba Bailian)
• Tavily API Docs
• Jina API Docs
• AMap MCP Docs

Test Cases

See DeepResearch.http for sample requests.

curl --location 'http://localhost:8080/chat/stream' \
--header 'Content-Type: application/json' \
--data '{
    "thread_id": "__default_",
    "enable_deepresearch": false,
    "query": "Please analyze the reasons for the explosive popularity of Pop Mart",
    "max_step_num": 2,
    "auto_accepted_plan": true
}'

📚 Reference Documentation

• Full Configuration Reference
• Spring AI Alibaba Documentation

🤝 Join Community & Contributing

Contributions are welcome! Please refer to CONTRIBUTING for guidelines.

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Contributors

Thanks to the following contributors for improving this project (unordered):

yingzi、zhouyou、NOBODY、xiaohai-78、VLSMB、disaster1-tesk、Allen Hu、Makoto、sixiyida、Gfangxin、AliciaHu、swl、huangzhen、Tfh-Yqf、anyin-xyz、zhou youkang、supermonkeyguys、yuluo-yx、Ken Liu、co63ox、benym

---