Introducing np — Deploy like K8s, Simple like Compose

Today we’re releasing np — a deployment platform for teams that need production-grade capabilities without production-grade complexity.

The problem

If you’re a small-to-medium team running 10-50 services on 3-15 servers, your options are:

• Docker Compose: Too simple. No health checks, no rolling updates, single-node only.
• Kubernetes: Too complex. Weeks to learn, months to stabilize, needs a dedicated ops person.
• Bare Nomad: Too bare. You have to assemble Consul, Traefik, monitoring, and logging yourself.

There’s a gap. And that gap is where np lives.

What np does

np bundles Nomad (scheduler), Consul (service discovery), Traefik (ingress), and OpenObserve (observability) into one platform. One binary to install. One command to deploy. One file to configure.

np init
np deploy my-app
np status

That’s it.

What makes np different

1. One file: No CRDs, no Helm charts, no Terraform. One np.yaml.
2. CLI-first: Every operation fits in 30 seconds.
3. Battery-included: You don’t assemble components. They come pre-integrated.
4. Runs anywhere: Raspberry Pi, old server, cloud VM. np treats them the same.
5. Honest about boundaries: np doesn’t try to replace K8s at scale. It’s for the 95% of teams that don’t need K8s.

Try it

curl -fsSL https://np.run/install.sh | bash
np init

np is MIT licensed. Core is free forever.

Get started →