A lightweight, self-hosted alternative to n8n/Zapier for internal automation.
Built with Redis queues, distributed workers, retries with exponential backoff, and a fault-tolerant execution engine.
OpsFlow is a backend-first workflow automation engine designed to solve real infrastructure problems such as:
- job orchestration
- distributed execution
- retries and backoff
- state persistence
- observability
This project intentionally avoids UI complexity and focuses on systems engineering.
- API (Next.js Control Plane)
- Triggers workflow execution
- Workflow Engine
- Resolves steps and dependencies
- Redis Queue (BullMQ)
- Buffers jobs and retries
- Distributed Workers
- Execute jobs in parallel
- Postgres
- Persists state and logs
-
API (Control Plane)
- Workflow creation
- Trigger management
- Execution inspection
-
Workflow Engine
- Step chaining
- Dependency resolution
- Execution state tracking
-
Redis Queue (BullMQ)
- Job buffering
- Retry scheduling
- Backoff management
-
Workers
- Horizontally scalable
- Stateless execution
- Idempotent job handling
-
Postgres
- Execution history
- Logs
- Workflow metadata
- ✅ Webhook + manual triggers
- ✅ Redis-backed job queue (BullMQ)
- ✅ Distributed worker processes
- ✅ Automatic retries + exponential backoff
- ✅ Step chaining
- ✅ Persistent execution history
- ✅ Dockerized infra (Postgres + Redis)
- ✅ Idempotent job design
This project prioritizes backend infrastructure challenges, including:
- job scheduling
- distributed systems
- failure handling
- retry strategies
- observability
- system resilience
cd docker
docker compose up
cd ..
pnpm --filter @opsflow/worker dev
pnpm --filter @opsflow/api dev
Most portfolio projects are CRUD apps. OpsFlow focuses on backend infrastructure problems:
- OpsFlow demonstrates hands-on experience with:
- real-world async workloads
- distributed execution models
- production-grade reliability patterns
- Built to scale — and to break — so it can be fixed properly.