Admin Operations Dashboard

Overview

The Admin Operations Dashboard gives platform administrators a real-time view of business metrics, system health, and performance data. Track MRR, subscriber growth, churn rates, and monitor system latency, data integrity, automated load test results, and AI autofix activity across Sentry errors, load tests, and CI workflows.

Accessing the Dashboard

1. Log in with an admin account
2. Navigate to /admin/operations/overview for business metrics
3. Navigate to /admin/operations/health for system health

Overview Page

Revenue Metrics

At the top of the dashboard, you'll see key financial indicators:

• MRR (Monthly Recurring Revenue) - Total subscription revenue across all active paid plans
• Active Subscribers - Number of companies on paid plans
• Churn Rate - Percentage of customers who canceled in the last 30 days
• Average Revenue Per User (ARPU) - MRR divided by active subscribers

Growth Metrics

Track platform growth over time:

• New Signups (Last 30 Days) - User and company registrations
• Trial Conversions - Percentage of trials that converted to paid
• Customer Lifetime Value (LTV) - Projected revenue per customer

Email Reputation

Monitor AWS SES sender health:

• Bounce Rate - Percentage of emails that bounced (14-day rolling window)
• Complaint Rate - Percentage of emails marked as spam
• Reputation Status - GREEN (healthy), YELLOW (warning), or RED (critical)
• Primary Recipient Sends - Count of warmup emails sent to nortonholdingsinc@gmail.com
• Secondary Recipient Sends - Count of warmup emails sent to chris.norton16@gmail.com

Why Email Reputation Matters: SES reputation is calculated as a 14-day rolling average. The warmup cron sends clean emails to verified seed inboxes every 3 minutes, pumping volume into the denominator to drive down bounce rates after fixing upstream issues like placeholder.local email skips.

Health Page

System Performance

Monitor technical health metrics:

• API Latency (P95) - 95th percentile response time for critical endpoints
• Database Query Time - Average query execution time
• Error Rate - Percentage of requests returning 5xx errors
• Uptime - System availability percentage over the last 24 hours

Data Integrity

Track data quality across the platform:

• Orphaned Records - Tickets without properties, jobs without tickets, etc.
• Invalid Addresses - Properties with missing or malformed geocodes
• Duplicate Customers - Potential duplicate customer records by email/name
• Missing Required Fields - Records with empty required data

Automated Load Test Results

View performance under simulated load:

• Last Run - Timestamp of most recent load test
• Result - PASSED or FAILED
• Virtual Users - Number of concurrent users simulated (500)
• Total Operations - Count of database operations executed
• Read Operations - Number of SELECT queries (90% of total)
• Write Operations - Number of INSERT/UPDATE queries (10% of total)
• Average Response Time - Mean time per operation
• P95 Response Time - 95th percentile response time
• Failures - Count of operations that errored or timed out

How Load Tests Work: Every Sunday at 12am UTC, BlueClerk simulates 500 virtual users performing mixed read/write operations. Tests are batched (50 concurrent users per batch) and automatically clean up all test data after completion. Detailed results are saved for historical comparison.

AI Autofix Activity

Track automated bug fixing across the platform:

• Sentry Errors Fixed - Number of production errors automatically resolved via GitHub Actions
• Load Test Failures Fixed - Automated fixes triggered by load test failures
• CI Workflow Fixes - Fixes deployed through the CI/CD pipeline
• Average Fix Time - Mean time from error detection to deployed fix
• Success Rate - Percentage of autofix attempts that resolved the issue

How Autofix Works: When Sentry detects a new error, load tests fail, or CI workflows break, GitHub Actions automatically triggers a Claude-powered autofix workflow. The AI analyzes the error, generates a fix, creates a branch, and opens a PR for human review. Successfully merged PRs increment the fix counters.

Using the Dashboard

Daily Monitoring

Check the dashboard daily to:

• Spot revenue trends and subscriber changes
• Monitor email reputation and warmup progress
• Identify system performance degradation
• Catch data integrity issues early
• Review load test results after weekly runs
• Track AI autofix effectiveness

Alerts and Thresholds

The dashboard highlights critical issues:

• Red badges - Immediate action required (bounce rate >5%, error rate >1%, failed load tests)
• Yellow badges - Warning state (approaching thresholds, degraded performance)
• Green badges - Healthy state (all metrics within normal ranges)

Historical Trends

Each metric includes a trend indicator:

• ↑ Green arrow - Improving metric (revenue up, errors down)
• ↓ Red arrow - Declining metric (churn up, uptime down)
• → Gray arrow - Stable metric (no significant change)

Questions

Q: How often does the dashboard update? A: Revenue and growth metrics update every 15 minutes. System health metrics update in real-time. Load test results update weekly after each Sunday run.

Q: What's considered a healthy bounce rate? A: AWS SES considers <5% healthy. 5-10% is a warning state. >10% risks account suspension. The warmup cron aims to keep bounce rates <3% by pumping clean email volume.

Q: Why do load test results only appear weekly? A: Load tests run every Sunday at 12am UTC to avoid impacting production traffic. Each test simulates 500 users over ~90 seconds and generates detailed performance data for the week.

Q: How can I investigate a specific metric? A: Click any metric card to drill into detailed logs, error traces, or historical graphs. Most metrics link to relevant admin tools or external monitoring dashboards.

Q: What triggers an AI autofix? A: Sentry errors with stack traces, load test failures with error logs, and CI workflow failures all trigger the autofix GitHub Actions workflow. The AI analyzes context and attempts a fix automatically.