edhenry / sregpt

Site Reliability Engineer Bot

Getting Started

Site Reliability GPT is a Large Language Model backed Site Reliability tool.

The goal of this project is to develop a tool for the Site Reliability Engineering community to use during the operations, administration, and maintenance of the systems and services they build and support.

Rough roadmap:

DevOps

• Local integration development environment
• Helm chart for K8s cluster deployment

Data Sources

• Ingestion Pipelines for popular SRE docs
  • Google SRE Handbook
  • Kubernetes Documentation
  • Docker Documentation
  • CNCF Project(s) Documentation
  • AWS EC2 Documentation
  • GCP Documentation

Core Functionality

• SREGPT ChatEngine
• In-band model evaluation pipelines
• Out-of-band labeling service

Integrations

• Kubernetes Events API Integration
• OpenShift Events API Integration
• OpenInference Observability Integration
• AWS CloudWatch API Integration
• Google Cloud Events API Integration