ReRun

Portfolio Project · Workflow Automation

Run it. Watch it. Safely run it again.

ReRun is an automation ops dashboard for small teams: run your workflows, see exactly what failed, and re-run them safely — without duplicating side effects. Observability and idempotent retries in one place.

See the demo → Why it matters
Runs
PlaybookTriggerStatus

01 — What it solves

When an automation breaks, you're flying blind.

Small teams rely on automations — Slack alerts, Sheets updates, emails. But when one breaks, there's no visibility and no safe way to re-run without doing the side effects twice. ReRun fixes both.

👁️

Observability

Every run, every step, every failure — visible in one dashboard instead of scattered logs.

🔁

Safe retries

Re-run failed workflows with idempotency, so retries don't duplicate side effects.

⚙️

Background jobs

Workflows execute asynchronously on a Redis + Celery queue, decoupled from the UI.

02 — Core model

Playbooks, steps, and runs.

A Playbook is a workflow definition made of ordered Steps. Each execution is a Run you can observe and replay.

Playbook
workflow definition
→
Steps
ordered actions
→
Run
one execution
↺
Re-run
idempotent retry

03 — Stack

Built with.

Backend

  • Python
  • FastAPI
  • PostgreSQL
  • SQLAlchemy + Alembic

Queue

  • Redis
  • Celery
  • Background jobs

Frontend

  • React
  • TypeScript
  • Vite
  • Tailwind · TanStack Query

Infra

  • Docker Compose
  • GitHub Actions CI

04 — Roadmap

From scaffold to demo.

WEEK 1

Core models + execution engine.

WEEK 2

Slack + Google Sheets integrations.

WEEK 3

Retries, idempotency, observability.

WEEK 4

Polish + demo.