Lakefront
Sign inStart free
← Lakefront

Deploys

Push to main.
That’s the deploy.

Connect a repository and Lakefront does the rest: it works out what your repo contains, builds an immutable image into your own registry, and rolls it out behind a health check. No pipeline to author, no Kubernetes to babysit.

Lakefront · acme
Platform
Services
Status
Recovery
Previews
Deploy
Clouds
Secrets
Databases
Domains
acme / Services acme/checkout Deploy
Checkout8 services · 2 clouds
checkout-apiwebeastusLive
checkout-webstaticeastusLive
ledger-workerworkerus-east-1Deploying
invoice-renderfunctionus-east-1Live
nightly-reconcilecronwesteuropeLive
search-indexerworkereu-west-1Live
webhooks-apiwebeastusLive
thumbnailerfunctionus-east-1Live

One definition, every provider

A service is described once. The same definition compiles to Azure Container Apps or AWS ECS and Lambda, with the same scaling, health and release semantics on both. Start on one cloud, add another next quarter, rewrite nothing.

  • Five workload kindsweb, worker, cron, static and function, each mapped to the right native primitive.
  • Framework presetsNext.js, Remix, Vite and plain static builds are detected and configured for you.
  • Monorepos understoodLakefront scans the repo, proposes each service it finds, and flags one nobody deployed.
lakefront.service.ts
// One definition. It compiles to each cloud's native primitives. export default service({ name: 'checkout-api', source: repo('acme/checkout'), workload: 'web', // worker · cron · static · function cloud: 'azure', // or aws regions: ['eastus', 'westeurope'], scale: { min: 1, max: 10 }, release: { strategy: 'gradual', drain: 30 }, });

Releases nobody notices

New revisions run alongside the old ones until they are healthy. Connections drain instead of dropping, lifecycle hooks run at the right moment, and any past revision is one click away on either cloud.

  • Overlapping and gradualShift traffic a slice at a time, or all at once when you would rather.
  • Health-gatedTraffic moves only after the new revision answers.
  • Instant rollbackReturn to a known-good revision without a rebuild.
ledger-worker · build #418
ledger-workerDeploying1m 06s
Clone
Detect
Build
Push
Release
Build logus-east-1
10:02:14 detected framework next.js 16
10:02:31 restoring build cache .next/cache hit
10:02:52 pushing ledger-worker:8f3c1a2 → your registry
10:03:07 rolling out revision 0042 · overlap 1
10:03:19 health check passing · shifting traffic

Priced before you commit

The deploy wizard estimates what a service will cost on the target cloud from that provider’s own rate card, before anything is created. After it ships, the same number is reconciled against your real bill.

Deploy
1Image2Environment3Review & deploy
Repository
acme/checkout · main
Service
checkout-api · web
Target
Azure · eastus
Scale
1 – 10 replicas
Environment
18 vars · 6 secret
Estimated costyour rate card
Container Apps · 0.5 vCPU$38 / mo
Container registry$5 / mo
Billed to your Azure subscription$43 / mo
BackDeploy

A real environment per pull request

Every branch gets its own URL, its own configuration and its own data boundary, torn down automatically when the pull request closes. Reviewers click a link instead of checking out a branch.

Previews
Open pull requestsauto-torn-down
#812 Split fares by riderpr-812.checkout.acme.devLive
#809 Retry failed uploadspr-809.checkout.acme.devLive
#806 Reduce ETA jitterpr-806.checkout.acme.devBuilding

Domains, DNS and TLS

Search and register a domain without leaving Lakefront, or bring one you already own. Records are written into Route 53 or Azure DNS, verified over DoH, and certificates are issued and renewed for you.

Domains · checkout.acme.com
checkout.acme.comVerified
A
checkout
20.51.14.8
Front Door
CNAME
www
checkout.acme.com
alias
TXT
_lakefront
lf-verify=8f3c…
ownership
TLS issuedRoute 53Azure DNSwildcard previews
  • Builds are cached in your own registry, so the second deploy is much faster than the first.
  • Every deploy is an immutable image; nothing is patched in place.
  • Cost is estimated before you deploy and reconciled against your real bill afterwards.

Keep reading

Agentsauth.md registration, attenuated scopes, and a model chain that falls back cleanly.ResilienceLive metrics, alert rules, multi-region failover, drills, status pages and incidents.SecurityOne permission model across every cloud, just-in-time elevation, one audit trail.DataManaged databases, object storage, and secrets that rotate themselves.

Deploy it into your own cloud.

Start freeBack to overview