Demo Walkthrough

Step 1: Create a Risky Change

Create a pull request that removes resource limits from a production deployment. This is a common mistake that can lead to resource exhaustion and cluster instability.

Step 2: PatchPulse Detects the Change

The Git integration automatically detects the PR and extracts the diff. The change is sent to the backend for analysis.

Step 3: AI Analysis Runs

PatchPulse analyzes the change using both rule-based guardrails and AI-powered analysis. It identifies:

• Missing resource limits (guardrail triggered)
• Security vulnerabilities (AI scan)
• Cost optimization opportunities (AI analysis)
• Incident prediction based on patterns (AI)

Step 4: Risk Score Generated

A risk score is calculated (e.g., 85/100) based on:

• Guardrail severities (60% weight)
• AI risk assessment (40% weight)
• Cluster context (current resource usage)

Step 5: Decision Made

In advisory mode, the change is logged with a detailed explanation. In enforce mode, high-risk changes (score ≥ 70) are blocked automatically.

Step 6: Notification Sent

A Slack notification is sent with:

• Risk score and summary
• Triggered guardrails with evidence
• AI recommendations
• Interactive approve/deny buttons

Step 7: View in Dashboard

All decisions are visible in the dashboard with full details, AI insights, and historical trends. You can search, filter, and export decision data.