Skip to main content
← Back to projects
Actual ORBIT dashboard in a sample environment with illustrative demo data 02 / POOYA LABS
PROJECT / 02Observability & monitoring

ORBIT

Stable v1.0.0

An engineering workspace for service health, incidents, and operational workflows across metrics, logs, and traces. Version 1.0.0 is released for the portfolio-demo scope, not as a hosted production service.

GoReactPostgreSQLObservability
Open demo ↗ View source ↗
UI /Interface captures
ORBIT dashboard on mobile
ORBIT dashboard on mobile
01 /The challenge

In distributed systems, service health, incidents, alerts, and metrics, logs, and traces are often scattered across tools. ORBIT brings these operational views into one cohesive workspace.

02 /Approach

A React dashboard connects to API and data models for incidents, alerts, deployments, SLOs, audit, metrics, logs, and traces. The public demo and local engineering stack are deliberately separate.

03 /Role & attribution

The README identifies Pooya as project owner and maintainer.

04 /Technical architecture

The Next.js dashboard is static on Pages. The interactive demo uses an Apps Script Web App, fixed-origin iframe, and postMessage/google.script.run bridge, storing synthetic data in a private spreadsheet. A separate local engineering stack uses Go, PostgreSQL, and an OpenTelemetry Collector with Prometheus/Loki/Tempo.

Two separate ORBIT paths
The public demo uses synthetic data; the telemetry engineering stack is a separate local path
05 /Implementation details
  • 01Ten views: Overview, Incidents, Alerts, Notifications, Deployments, SLOs, Audit, Metrics, Logs, and Traces.
  • 02Incident lifecycle/timeline, bounded alert rules, and HTTP-success SLOs.
  • 03Bounded correlated telemetry search; no arbitrary PromQL/LogQL.
  • 04Persian/English, RTL, light/dark, responsive, keyboard, and chart table alternatives.
  • 05CI verifies the local OTLP path and workspace A/B isolation.
06 /Engineering decisions
  • 01The public demo uses synthetic data only; notifications are simulated and send no real messages.
  • 02Pages serves only the frontend; Go/Postgres/Prometheus/Loki/Tempo are not hosted production services.
  • 03Redis and NATS are in Compose but application integration is deferred.
  • 04v1.0.0 stabilizes the portfolio-demo scope, not a production SaaS.
07 /Verification & evidence
  • 01Stable v1.0.0 is published; release workflow, CI, Security, and Pages succeeded on the release commit.
  • 02Release CI passed 4/4 jobs and Security passed 13/13; Playwright/axe and NVDA evidence is recorded.
  • 03Live Apps Script bootstrap/mutation/refresh persistence need owner smoke testing; TalkBack/VoiceOver and full human review remain open.
09 /Capabilities & highlights
  • 01Incident lifecycle with an immutable timeline
  • 02Bounded alert-rule evaluation and SLO workflows
  • 03Bounded log and trace search with correlation
  • 04Persian/English dashboard with accessibility considerations
  • 05Stable v1.0.0 release for the portfolio-demo scope
10 /Scope & boundaries

The public demo uses synthetic data and private Google Sheets storage; the local engineering stack uses PostgreSQL and telemetry tooling. The v1.0.0 release does not mean every component is hosted as a production service.