pgpulse pgpulse
Get Started
Postgres-native observability

Observe every Postgres. Scale with confidence.

Intelligent Postgres observability with Pulse scoring, alerts, advisory, AI-powered runbooks, and hosted MCP access for AI workflows, turning database signals into clear visibility, guided investigation, and confident action.

Get Started Schedule a Demo
app.pgpulse.io
LIVE
Pulse Score Dashboard
Pulse Score Dashboard
WAL Pipeline Deep Dive
WAL Pipeline Deep Dive
Alerts
Alerts
Critical Gates
Critical Gates
Golden Signals
Golden Signals
Grafana Deepdive
Grafana Deepdive
<60s Time to first insight
24/7 Health checks
99.9% Platform uptime SLA
28-day Metric retention

Works with every Postgres provider

Supabase Supabase
AWS RDS AWS RDS
AWS Aurora AWS Aurora
GCP Cloud SQL GCP Cloud SQL
Azure PostgreSQL Azure PostgreSQL
DigitalOcean DigitalOcean
Neon Neon
Timescale Timescale
Aiven Aiven
Crunchy Bridge Crunchy Bridge
Heroku Postgres Heroku Postgres
Render Render
Railway Railway
Fly.io Fly.io
CockroachDB CockroachDB
Self-Hosted Self-Hosted
Supabase Supabase
AWS RDS AWS RDS
AWS Aurora AWS Aurora
GCP Cloud SQL GCP Cloud SQL
Azure PostgreSQL Azure PostgreSQL
DigitalOcean DigitalOcean
Neon Neon
Timescale Timescale
Aiven Aiven
Crunchy Bridge Crunchy Bridge
Heroku Postgres Heroku Postgres
Render Render
Railway Railway
Fly.io Fly.io
CockroachDB CockroachDB
Self-Hosted Self-Hosted
DB Pulse Scoring System

Your database health,
expressed as a single score.

pgpulse continuously evaluates 11 health domains — from Freeze Risk to WAL Pipeline — and rolls them into a single 0–100 Pulse Score. When a gate fires, you know exactly what broke and how to fix it.

11 Health Domains,
Scored & Sorted by Impact

Every domain is scored independently on a 0–100 scale, weighted by criticality, and sorted so the worst-performing areas always surface first.

  • Freeze Risk, Replication, Connections, Locks, Bloat, Vacuum, Queries, WAL, Disk, Objects & Memory
  • Weighted scoring — high-impact domains count more toward Pulse
  • Sorted by current score so your weakest area is always visible first
  • Per-domain drill-down with contributing raw metrics
Explore health domains
app.pgpulse.io
pgpulse health domains

Product

Observability built for every Postgres team

View product overview

Dashboards, alerts, and database insights — built for teams, no setup required.

Monitoring

Postgres metrics.
Zero config.

Built directly on native Postgres metrics, pgpulse renders dashboards instantly — connections, latency, error rates, memory, and storage. No agents. No sidecars. No extra stack.

  • 1-click import → live dashboards in under 60 seconds
  • Golden Signals across all your projects
  • 1h / 24h / 7d time ranges with drill-down
  • Multi-project rollup view
Start monitoring free
app.pgpulse.io
pgpulse golden signals
app.pgpulse.io
pgpulse alerts
Alerts

Alerts that follow
your metrics.

Turn any Postgres metric or advisory finding into a smart alert. Route to Slack, email, or webhook — grouped by project, with noise controls and full alert history.

  • Threshold and rate-of-change alert rules
  • Prebuilt and custom alert templates
  • Slack, webhook, and email routing
  • Alert history and noise suppression controls
Set up your first alert
Runbooks

AI-powered runbooks
for real incidents.

pgpulse turns full-instance state into pinned investigation guides that help operators understand what is happening, why it matters, what to validate first, and which next steps are safest.

  • Generated from full-instance state, Pulse, gates, alerts, telemetry freshness, and domain evidence
  • Structured into timeline, observations, root cause, validation, and recovery guidance
  • Preserves each runbook as a pinned, versioned incident artifact
  • Helps teams respond faster while keeping human judgment in control
See runbooks in action
app.pgpulse.io/runbooks
pinned
Runbook overview with incident summary, urgency, blast radius, and telemetry context
Runbook incident timeline showing cause, intermediate symptoms, impact, and prior context
Runbook key observations with evidence-backed findings and severity labels
Runbook root cause analysis with confidence level and alternative hypothesis guidance
Runbook action guidance showing safe next steps, undo path, and escalation boundary
Incident Snapshot
Hosted MCP

Use the pgpulse MCP server
with your LLM.

Connect external AI clients to pgpulse through a hosted MCP server so they can investigate health posture, domain evidence, gates, alerts, and recovery context using pgpulse’s investigation model instead of raw database access.

  • Read-only MCP access built for investigation, triage, and evidence retrieval
  • Token-based setup for external LLM clients and operator agents
  • Exposes high-signal tools, resources, and prompts for live PostgreSQL diagnosis
  • Safe by default without direct writes or raw SQL execution through the MCP layer
Set up MCP access
app.pgpulse.io/mcp
read-only
install hosted mcp prompt in cli read-only
claude / codex / cursor
$ claude mcp add pgpulse https://app.pgpulse.io/api/v1/mcp --header "Authorization: Bearer pmcp_live_..."
> list my postgres instances
• pgpulse-dev-0
instance_id: 557ecaea-10a9-4e78-a505-2abc9e57237d
status: active
collector_mode: self_hosted
health: watch
open_alerts: 2
> how is my instance health
pgpulse-dev-0 is currently watch as of 2026-05-21 10:21:42 +02:00.
The issue is not systemic right now: there are 0 active gates, but 2 open warning alerts.
The degraded domains are bloat and vacuum_engine.
Strongest signals: dead tuples at 14.11%, autovacuum saturation at 92.79%.
> how is my instance bloat and index health ready to continue
Advisory

Built-in database
health checks.

pgpulse runs a continuous advisory engine over your Postgres — surfacing slow queries, index bloat, vacuum lag, and more. Every finding comes with ready-to-run SQL and a clear severity level.

  • Index bloat, unused indexes, and missing index detection
  • Vacuum and autovacuum health monitoring
  • Query performance analysis with ready-to-run SQL
  • Promote any advisory finding directly to an alert rule
Run an advisory check
app.pgpulse.io
pgpulse score history with active gate details

Workspace

pgpulse org

Invite
L

Lena

Admin · joined 2d ago

Admin
R

Ryan

Maintainer · alerts

Maintainer
A

Amir

Viewer · metrics

Viewer

Shared: #infra-supabase, #db-alerts

Team

Invite your team.
Share the signal.

Share dashboards with devs, route alerts to ops, keep database advisory visible to platform teams. Simple roles, shared alert channels, and a full audit log.

  • Role-based access: Admin, Maintainer, Viewer
  • Shared alert targets per project
  • Per-project and cross-project views
  • Audit log for all configuration changes
Invite your team

More from the product

Main Dashboard
Main Dashboard
System Overview
System Overview
Memory Fit
Memory Fit
Grafana Explore
Grafana Explore
Query Performance
Query Performance
Alert History
Alert History
Score History & Gates
Score History & Gates

How it Works

From your DB to insights in minutes

Connect your Postgres securely — pgpulse instantly transforms your data into live dashboards, alerts, and insights.

1

Select Provider

Choose Supabase, AWS, GCP, Azure, or any supported Postgres provider.

2

Connect

Link your project securely with OAuth or a connection string.

3

Observe & Optimize

Dashboards and alerts light up with native metrics in under a minute.

Pricing

Simple per-project pricing, no surprises

Start free, scale as you grow. Full access from day one.

14-day free trial · No card required

Basic

Essential Postgres monitoring for every Supabase project.

$ 50 / DB / mo
  • Pulse Monitoring
  • AI Recommendations
  • Index Advisor
  • Vacuum & Bloat Analysis
  • Query Performance Insights
  • 7-day metric history
  • Up to 3 team members
Start free trial
Most Popular

Pro

Real-time signals, alerts, and deeper insights for growing teams.

$ 65 / DB / mo
  • Everything in Basic
  • Real-time Golden Signals
  • Detailed Metric Dashboards
  • Slack & Email Alerts
  • 30-day metric history
  • Up to 10 team members
Start free trial

Need custom retention or more team members? Talk to us about Enterprise.

Free 14-day trial

Your Postgres deserves better visibility

Connect your first database in under 60 seconds. No credit card required. Full access from day one.

Start for free — no card needed
Schedule a demo instead
No credit card 14-day free trial Cancel anytime