Platform Observability

How We Monitor Avsentia

We monitor every layer of our platform 24/7 — from database query performance to AI agent response latency — so you can focus on your BC/DR planning, not ours.

Monitoring Coverage

Six independent monitoring layers provide full-stack visibility across the platform.

Application Performance

Sentry

Real-time error tracking and stack traces
Performance profiling per endpoint
Alert on P95 response time > 2s
Release health tracking per deployment

Infrastructure & Uptime

Vercel

Edge network health across 30+ regions
Deployment rollback on build failure
Request volume and 4xx/5xx rate monitoring
CDN cache hit ratio and latency metrics

Database & Storage

Supabase

Query performance and slow query detection
Connection pool utilization
Storage capacity and I/O throughput
Automated daily point-in-time backups

External Availability

External Probes

HTTP endpoint probes every 60 seconds
SSL certificate expiry monitoring (alert at 30 days)
DNS resolution health
Multi-region availability checks

Security & Audit Events

Supabase Audit Logs

Failed authentication spike detection
Unusual query patterns or high-volume scans
Admin action audit trail
Subscription and payment anomaly alerts

AI Agent (Dave)

Custom Telemetry

Session success and error rate
LLM response latency (P50 / P95 / P99)
Token usage and quota tracking
Speech-to-text and TTS failure rate

Alerting and Response

When something goes wrong, the right person is paged within minutes.

Alert Channel	Trigger Condition	Response Time
Email	All SEV 1 and SEV 2 incidents	< 5 min
On-call rotation	SEV 1 (critical) incidents	< 15 min
Sentry alert	Error rate spike > baseline + 3σ	Automated
Vercel webhook	Deployment failure	Automated

Service Level Commitments

Our uptime and support SLAs by subscription tier.

Plan	Uptime SLA	Support Response	Coverage Window
Starter	99.5%	48 hours	Business hours
Professional	99.9%	12 hours	Extended hours
Enterprise	99.95%	4 hours	24 / 7 / 365

SLA credits apply if monthly uptime falls below the committed threshold. See Terms of Service for calculation details.

Backup and Recovery

< 1 hour

RPO

Recovery Point Objective — maximum data loss window

< 4 hours

RTO

Recovery Time Objective — maximum downtime per incident

Daily

Backup Frequency

Point-in-time recovery with 7-day rolling backup window

Questions about our platform reliability or SLAs?

support@avsentia.com Live Status Security