AI Status Dashboard

What this does

AI Status Dashboard answers one urgent question fast: is this AI outage affecting everyone, or is it just my setup? The product tracks major providers, translates live reliability signals into plain language, and gives people a decision in seconds instead of forcing them to parse raw status feeds.

The core experience is intentionally split: Casual Mode is optimized for fast, low-friction answers during incident stress, while the developer surface exposes machine-readable status for integrations, automation, and routing policy decisions. Both views are generated from the same underlying evidence.

Why this exists

When ChatGPT, Claude, or Gemini fail, users usually cannot tell whether the issue is local, regional, account-specific, or provider-wide. Official provider pages can lag, and generic monitoring dashboards often miss AI-specific surfaces like model routing, tool calls, and conversational latency. Independent, AI-focused monitoring closes that gap and gives users a practical “what now?” path during incidents.

The project started from repeated real-world support situations where teams lost time debugging their own infrastructure during provider-side incidents. A narrow AI reliability lens reduces that confusion: users see current symptoms, expected resolution patterns, and fallback paths in one place.

Who this is for

• The panicked user: quickly answers “is it just me?” and provides immediate fallback options.
• The SRE or engineer: exposes reliability data, incidents, and status trends for production decisions.
• The AI agent: publishes machine-readable status via API, OpenAPI, datasets, and MCP for automated routing.

How it works

AI Status Dashboard combines official status pages, observed reliability metrics, synthetic checks, and crowd signals into a single evidence-backed response. Casual Mode turns that evidence into clear, actionable copy for humans. Developer routes expose the same underlying signals for automation.

For deeper methodology details, evidence contracts, and response schema, use the public docs and API spec.

Every public route is designed for direct ingestion: OpenAPI for typed clients, llms.txt for language-model discovery, datasets for independent verification, and MCP for tool-based agent workflows. The goal is to make reliability claims auditable, not just readable.

How we’re different

This is not a generic status aggregator. It is purpose-built for AI providers, returns evidence-backed responses with confidence scoring, and publishes public datasets plus API and MCP surfaces for programmatic use. That combination lets both humans and systems validate claims and act quickly.

In practical terms: the product focuses on the AI reliability questions people actually ask under pressure, keeps machine-readable surfaces synchronized with user-facing pages, and exposes historical context such as typical resolution windows. It is built to support both immediate triage and long-term reliability analysis.

Project status

AI Status Dashboard is built by Khalid Saidi, open-source under MIT, and maintained as part of a broader portfolio of agent-readiness tools. Releases and maintenance history are published in the changelog.

Questions or bugs: open an issue in the GitHub repository.