Point-in-time restore is now on every plan

Ship containers and databases without the ops

Push code and get a built image, a health-checked rollout, managed Postgres and continuous backups — on infrastructure you never have to log into.

Deploy your first serviceSee how it works
$npx infraline deploy
scheduler · us-west · live
Scheduler60stick · dispatch · ackweb-1*/1 * * * *worker-1*/1 * * * *cron-1*/1 * * * *1,440 runs/day · 0 missed0s downtime
median build
21s

median build

regions
5

regions

uptime SLA
99.99%

uptime SLA

YAML files
0

YAML files

Trusted by teams shipping every day

  • Northwind
  • Vantablack
  • Orbital
  • Latchkey
  • Fernweh
  • Hexline
  • Quanta
  • Driftwood
  • Northwind
  • Vantablack
  • Orbital
  • Latchkey
  • Fernweh
  • Hexline
  • Quanta
  • Driftwood

One platform

Every piece of your stack, deployed the same way

Services, containers, databases and jobs share one build system, one private network and one set of secrets. Nothing to wire together.

Web Services

Auto-detected buildpacks, zero-downtime rollouts, instant rollbacks and preview URLs per pull request.

Node · Python · Go · Ruby · Rust

Containers

Bring any Dockerfile. We build it, scan it, and schedule it across the fleet with health-checked restarts.

OCI · Dockerfile · registry sync

Managed Postgres

Provisioned in seconds with connection pooling, read replicas and encrypted storage by default.

Postgres 17 · pgvector · PgBouncer

Key-Value & Cache

Redis-compatible instances with persistence, eviction policies and private networking to your services.

Redis 7 · AOF · TLS

Cron & Workers

Background jobs and scheduled tasks that share your service image, secrets and private network.

cron · queues · one-off jobs

Static & Edge

Immutable static deploys pushed to every edge PoP with atomic invalidation on each release.

SPA · SSG · edge redirects

Automation

From git push to serving traffic, untouched by human hands

Four stages, all of them ours to run. You get a URL, a log stream and a rollback button.

git pushmain@8f3c1d2Buildcache hit · 21sweb-1liveweb-2liveweb-3livepostgres

  1. 01Push

    A commit, a tag or a merged PR. We watch the branch you pick and nothing else.

  2. 02Build

    Runtime auto-detected or your Dockerfile honoured, built in a clean sandbox with a warm layer cache.

  3. 03Verify

    Your checks run, the image is scanned for CVEs, and secrets are injected only at boot.

  4. 04Release

    Health-checked rollout, traffic shifted when the new instances answer, one-click rollback if they don't.

build.logdeployed

==>Cloning repo at 8f3c1d2

==>Detected Dockerfile — building image

ok Layer cache hit (7/9 layers)

ok Image pushed sha256:4b9a…c17e

==>Running health check GET /healthz

ok 3/3 instances healthy — shifting traffic

==>Live at api-gateway.infraline.app (21.4s)

Always on

Cron that fires every minute, on containers that never go down

One scheduler, one clock, every minute of the day. Your containers stay warm, the dispatch is acknowledged, and nothing quietly stops running at 3am.

A tick every 60 seconds

The scheduler runs on its own clock, not inside your app. If a container is mid-restart when the minute lands, the run is queued, not skipped.

Missed runs replay themselves

Every dispatch is acknowledged. Anything unacknowledged inside the window is retried with backoff until it completes or you cancel it.

Deploys don't stop the clock

New instances take traffic only after they answer a health check, so a release never lands between two ticks and drops one.

Databases & backups

Backups that are a timeline, not a cron job

Every managed database streams its changes continuously and snapshots daily. Recovery is a timestamp you pick, not an archive you hope someone wrote.

primaryrestored-6d-5d-4d-3d-2d-1fork readyrestore 14:32:07 UTC35-day retention windownow

Point-in-time restore

Write-ahead logs ship continuously, so you can rewind to any second inside your retention window — not just to last night's dump.

Instant forks

Branch production into an isolated copy in seconds. Test the migration, throw the fork away, keep the data.

Encrypted end to end

AES-256 at rest, TLS in transit, keys rotated automatically and never shared between tenants.

Your bucket, your copy

Mirror every snapshot to an S3-compatible bucket you own. No lock-in, no export tickets.

Global network

Close to your users, private between your services

Deploy to five regions with private networking inside each one. Services talk over an encrypted mesh that never touches the public internet.

  • Oregonus-west
    18 ms
  • Virginiaus-east
    22 ms
  • Frankfurteu-central
    26 ms
  • Singaporeap-southeast
    31 ms
  • São Paulosa-east
    34 ms

Pricing

Priced per service, not per surprise

Compute and storage billed by the second. Backups, TLS, preview environments and private networking are included on every plan.

Hobby

For side projects and prototypes.

$0/ month

  • 1 web service + 1 container
  • 512 MB Postgres, daily backups
  • Automatic HTTPS & preview deploys
  • Community support
Start free
Most popular

Pro

For production apps that need to stay up.

$23/ month

  • Unlimited services & containers
  • Point-in-time restore (7 days)
  • Autoscaling & zero-downtime deploys
  • Private networking + custom domains
  • Email & chat support
Start 14-day trial

Scale

For teams with compliance and SLOs.

$82/ month

  • Dedicated compute & isolated VPC
  • Point-in-time restore (35 days)
  • Cross-region backup replication
  • SSO/SAML, audit logs, RBAC
  • 99.99% uptime SLA
Talk to sales

FAQ

Questions worth asking before you migrate

You keep the same primitives — containers, private networks, scheduled jobs — without owning the control plane, the upgrade cycle, or the on-call rotation that comes with them.

Your first deploy takes about four minutes

Connect a repo, pick a region, and watch it go live. No credit card, no cluster, no YAML.

Start freeRead the docs