Implement Monitoring & Observability (APM, Metrics, Alerting)

[carlos-israelj]

Description

No monitoring infrastructure. Cannot track performance, errors, or system health. No alerts when service degrades.

Current State

Basic Winston logging only.

Problems

• No error tracking
• No performance monitoring
• No alerting
• Cannot debug production issues

Proposed Solution

Full observability stack: Sentry (errors), Prometheus/Grafana (metrics), alerts.

Acceptance Criteria

• Sentry integrated (backend + frontend)
• Prometheus metrics exported
• Grafana dashboards (response time, error rate, throughput)
• PagerDuty/Slack alerts for critical errors
• Distributed tracing (OpenTelemetry)
• Performance budgets enforced

ETA

3-4 weeks