OpenDeRisk: A 24/7 AI Digital Operations Assistant for Application Systems
What is OpenDeRisk?
OpenDeRisk is a 7×24 risk intelligence system, like a digital operations assistant that can help application systems solve problems at any time and work alongside people. We hope every application system can have such a digital steward, guarding system security at all times.

OpenDeRisk Project Features
Deep Research Root Cause Analysis (DeepResearch RCA): Quickly identifies the root cause of issues by deeply analyzing logs, traces, and code.
Visual Evidence Chain: Comprehensively presents the diagnostic process and evidence chain, making diagnostic results clear and easy to understand.
Multi-Agent Collaboration: Collaboration among SRE-Agent, Code-Agent, ReportAgent, Vis-Agent, and Data-Agent to jointly perform risk analysis and diagnosis.
Open-Source Architecture: OpenDeRisk's architecture is fully open, making it easy to use and integrate in open-source projects.
OpenDeRisk Technical Implementation
Data Layer: Pulls a large-scale OpenRCA dataset (approximately 20GB) from GitHub and performs local processing and analysis on it.
Logic Layer: Adopts a multi-agent architecture, where collaboration among agents such as SRE-Agent and Code-Agent jointly completes deep root cause analysis.
Visualization Layer: Uses the Vis protocol to dynamically render the entire processing workflow and evidence chain.
Installation and Configuration
Install uv:
curl -LsSf https://astral.sh/uv/install.sh | sh
Install dependency packages:
uv sync --all-packages --frozen \ --extra "base" \ --extra "proxy_openai" \ --extra "rag" \ --extra "storage_chromadb" \ --extra "client" \ --index-url=https://pypi.tuna.tsinghua.edu.cn/simple
Configure the API_KEY in the derisk-proxy-deepseek.toml file, then run the following command to start:
uv run python packages/derisk-app/src/derisk_app/derisk_server.py --config configs/derisk-proxy-deepseek.toml
Visit http://localhost:7777 to view the running results.

OpenDeRisk Advantages
Comprehensive Open-Source Architecture: Provides complete open-source code and architecture, allowing users to freely customize and integrate.
Application of Deep Learning: Deeply applies deep learning technology, improving the accuracy of root cause analysis.
Visual User Experience: Through visualization, the diagnostic process becomes more intuitive, easy to understand, and operate.
Community Support: Has an active community, providing good technical support and a user communication environment.
Related Links
https://github.com/derisk-ai/OpenDerisk
Script Planner Assistant▸
AI Apps
AI News 