PulseBoard — Health dashboard for your AI endpoints

See at a glance whether the AI proxies you rely on are actually usable

Problem

Developers using reverse proxies or relay services to reach OpenAI-compatible models often hit slow responses or rate limits and can't tell network issues from flagged accounts or upstream outages. Communities just ask each other 'is it acting up again?'

Solution

Periodically probe multiple OpenAI-compatible endpoints, sample latency and error codes, keep history, and classify incidents as upstream outage, rate limit, or local network—with email alerts on state changes.

Users

Indie developers and small teams relying on relay APIs, AI app operators

MVP

Add multiple compatible endpoints and keys; probe every 5 minutes for latency and status; historical trends; email or webhook alerts; shareable public status page

Revenue

Free for 2 endpoints; subscription for more endpoints, longer history and team alerts

Why now

Many developers reach models through proxies and relays where upstream churn is constant, and generic monitoring tools don't understand OpenAI-compatible semantics—a vertical tool is cheap to build.