AI DevOps Agent — Autonomous Risk-Aware CI/CD

AI DevOps Agent is an intelligent DevOps automation system that analyzes every GitHub commit before it reaches deployment.

The problem is that traditional CI/CD pipelines primarily answer "Can this code be built?" but often do not answer "Is this change safe to deploy?" Developers may miss security-sensitive changes, exposed secrets, risky configuration changes, or potentially dangerous code modifications.

Our solution creates an automated commit-to-deployment decision pipeline.

When a developer pushes code to GitHub, a webhook sends the event to our FastAPI backend. The system retrieves the commit and relevant changes through Swytchcode, which provides a controlled execution layer for API integrations.

The AI analyzer evaluates the change across multiple risk factors and produces a risk score from 0–100.

The Decision Engine then determines the appropriate action:

• LOW/MEDIUM risk → proceed toward automated deployment through Netlify
• HIGH/CRITICAL risk → block deployment and create a Jira incident for developer review
• Security/secret detection can override the normal deployment decision

The complete workflow is:

Developer → Git Push → GitHub Webhook → FastAPI → Swytchcode → GitHub API → AI Risk Analysis → Decision Engine → Netlify Deployment OR Jira Incident

Swytchcode is used as the execution authority between our AI agent and external APIs. Instead of allowing the agent to directly call GitHub, Jira, or Netlify APIs, API operations are executed through defined Swytchcode tools with structured inputs, authentication, validation, and controlled execution.

The project also provides a dashboard showing commit analysis, risk scores, deployments, incidents, analytics, and AI diagnostics.

This project demonstrates how AI can be integrated into DevOps not simply as a chatbot, but as an automated decision-making layer inside a real software delivery workflow.