Database monitoring should lead to an informed next step.

One portfolio, distinct sources

Connect Neon and Supabase Management APIs, add compatible PostgreSQL databases directly, and review the resulting projects together. Provider terminology, periods, units, and provenance remain attached to every measurement.

• Portfolio-wide size, growth, status, and synchronization views
• Project timelines and provider-aware metric labels
• Missing values remain unavailable instead of becoming zero

Provider monitoring that respects boundaries

DatabaseMeter collects only capabilities exposed by each supported provider and credential scope. Neon consumption, endpoint configuration, Supabase project usage, and provider resource evidence remain separate when they are not directly comparable.

• Project discovery and selective monitoring
• Provider usage and resource evidence where available
• A published provider capability matrix

Optional PostgreSQL telemetry

A dedicated read-only role can add database, relation, session, and aggregate workload context to a provider project or monitor a direct PostgreSQL database. Collection can run manually or on a per-database schedule.

• Database, table, and index size observations
• Connection, cache, I/O, maintenance, and activity statistics
• pg_stat_statements fingerprints and counters without SQL text

Explain, do not merely graph

The workspace is organized around practical questions: what changed, what is growing, what needs attention, and what evidence might explain a workload. Deterministic rules and explicit caveats keep those answers reviewable.

• Largest and fastest-growing databases
• Provider compute kept separate from observed SQL workload
• Qualified projections and findings rather than fabricated certainty

Optional, on-demand AI interpretation

Authenticated users can choose a predefined operational question and analyze one database, all monitored databases, all Neon, all Supabase, or all other monitored sources. DatabaseMeter assembles only the prompt-relevant normalized evidence, validates the structured result, and retains it for 30 days unless the user deletes it sooner.

• Evidence-linked findings, context, confidence, and reversible next steps
• Credentials, connection details, SQL text, application rows, and raw provider responses excluded
• Read-only, rate-limited requests that never execute a suggested change

Alerts, history, and reporting

Stored observations support trend views, in-app alert history, synchronization diagnostics, and optional weekly reports. Rules evaluate only when the required metric is actually available.

• User-configured thresholds and explainable attention states
• Recorded collection and delivery outcomes
• Opt-in weekly summaries of noteworthy changes

Useful before connecting anything

Public calculators, PostgreSQL inspection tools, provider comparisons, and the Learn Center help visitors investigate a result or understand a metric without first creating an account.

• Local health-check and pg_stat_statements analysis
• Growth, CU-hour, egress, and usage calculators
• Source-backed PostgreSQL guides with interpretation guardrails

Observe infrastructure, not customer data

Credentials are validated and encrypted server-side. PostgreSQL collection uses bounded statistics queries and excludes application rows and SQL text. Direct destinations require verified TLS and server-side network validation.

• Server-enforced ownership checks
• Saved secrets are never displayed again
• Disconnect and permanent DatabaseMeter account-data deletion controls

Built by AI Sure Tech

DatabaseMeter is an AI Sure Tech project focused on practical database monitoring and developer infrastructure. Accounts and dashboard access are free to create and use, with no credit card or payment information required. Deterministic calculations, provider APIs, and PostgreSQL statistics remain the core workspace; optional AI analysis is user-triggered and advisory.