Hostess

Platform

Why teams choose Hostess.

Hostess occupies a specific gap: Compose-like developer experience with Kubernetes-grade multi-service semantics — without forcing teams into raw cloud primitives or dashboard-first workflows.

Big Cloud

AWS · GCP · Azure

  • 15+ services just to deploy a simple app
  • IAM roles, VPCs, security groups, permission boundaries
  • Forgotten resources rack up bills while you sleep
  • One misconfigured permission breaks everything at 2 AM

Managed PaaS

Vercel · Railway · Render

  • Platform-specific config models — portability tax on exit
  • State split between code, dashboard, and env overrides
  • Multi-service apps require service-by-service setup
  • Switching providers means rewriting everything

DIY Kubernetes

k8s · Helm · Terraform

  • Clusters, ingress, operators, secrets, observability
  • 450+ lines of YAML to deploy a simple web app
  • Deployment knowledge siloed to infra experts
  • High ops burden with fragile team handoffs
AI-agent native

The platform AI agents would deploy to themselves.

Every AI-powered app is a multi-service stack. Hostess's entire workflow is CLI and config-file driven — after a one-time setup, an autonomous agent never needs to touch a GUI again.

  • No browser automation or computer-use overhead
  • Pure stdin/stdout — native to any LLM agent loop
  • Write code → deploy → read logs → fix → redeploy
  • The full loop stays in the terminal where agents are fastest
agent loop
# authenticate once via env var
export HOSTESS_TOKEN=pat_xxxxxxxx
# the entire loop — no GUI needed
$ hostess deploy
✓ stack deployed in 42s
$ hostess services logs api
ERROR: missing env var OPENAI_KEY
$ hostess secrets add OPENAI_KEY
✓ secret stored
$ hostess deploy -s api
✓ api redeployed in 18s

Built for every layer of your stack.

Operator-backed databases

Postgres via CloudNativePG. Redis with persistence. Automatic backups, connection pooling, HA.

Magic variables

Internal URLs, external URLs, connection strings. Resolved automatically. Never copy-paste a database URL again.

Lifecycle hooks

Run migrations before traffic switches. Seed databases on first deploy. Warm caches after deploy.

Deployment URLs

Every deploy gets a unique SHA-based URL. Stable routes update when the new deploy is healthy.

Per-service deploys

Target specific services with hostess deploy -s api. Fast iteration on a live stack.

AI-agent native

Full CLI-driven workflow. The deploy–debug–redeploy loop stays in the terminal where agents are fastest.

Ready to ship your stack?

Free to start. No Kubernetes expertise required.