Solutions · Applications

Room to run.
Rails underneath.

SaaS, APIs, and the services behind them — on NVMe-backed nodes kept deliberately under-filled, with your database included and a protection posture that can restore to the minute. Platform-grade iron, without a platform’s worth of ceremony.

GET/api/orders200·84 ms total
dns
11ms
tls
23ms
app
33ms
db · nvme
9ms
response
17ms

db on local NVMe — no network hop between your app and its data. Low-density nodes keep the numbers looking like this at 4 p.m. too.

representative trace · product art

The platform

What your app stands on

01

NVMe end to end

Application and database share local NVMe — no network storage hop in your hot path. Reads come back in microseconds, and p99s stop telling stories.

02

Low-density by policy

Nodes are kept deliberately under-filled, so your neighbors cannot become your latency. When the fleet nears its engineered limits, we add iron — we never oversell.

03

Databases included

Every tier ships with a managed database allowance — 1, 10, or 100 GB by plan — minded, backed up, and on the same node as your app.

04

Watched like production

Checks every sixty seconds, 24/7, with same-day answers from people who can see your node — because it is ours.

Deploy anything

Ship it the shape
it already is

No house framework, no proprietary build step, no rewrite tax. If it serves HTTP, it can take a berth — and it inherits the fleet’s monitoring, backups, and TLS the moment it lands.

Static & prerendered

Built output served fast with sane caching headers. The boring case, done properly.

html · assets · frameworks that export

Node services

Long-running services with process supervision, log capture, and restarts that do not need you.

apis · ssr · workers

Containers

Bring an image; it gets a berth, resource bounds, and the same monitoring as everything else on the node.

anything that builds

Continuous Protection

When the bad deploy lands,
how much do you lose?

Continuous Protection is a protection feature, not a different kind of hosting: live replication streams every write to a second node within seconds, and a change journal lets us restore your data to a chosen minute — not to whenever last night’s snapshot ran. Included in Foundry; available as an add-on on Ember and Forge.

One day · one incident · two outcomes

schematic — not a dashboard

Nightly snapshot

every plan, included

02:00
restore rolls back to 02:00 — ≈14 h of changes gone

Continuous Protection

included in Foundry · add-on elsewhere

every write replicated · journal kept
restores to 16:06:41 — the minute before · loses seconds
00:0006:0012:0018:0024:00

16:07 — a bad import overwrites live data. Both rows face the same moment; only the posture differs.

continuous protection — live replication + point-in-time restore · protection feature, shown as schematic

Databases

Included, minded,
and next to your app

Every plan includes a managed database allowance on the same NVMe as your application. Backed up nightly with the rest of the site, watched with the rest of the node — no separate database bill, no surprise storage tier.

Outgrowing a cap is a plan change, not a migration.

PlanDB includedPrice

Ember

side projects, small app backends

1 GB

$19/mo

Forge

production apps with real users

10 GB

$39/mo

Foundry

data-heavy apps · Continuous Protection included

100 GB

$79/mo

Enterprise

Odd shape? Custom needs? That is a conversation.

Dedicated capacity, custom protection windows, unusual runtimes, compliance questions — Enterprise berths are specified with you, priced plainly, and run on the same fleet discipline as everything else.

Talk to us

For applications

Put your app on
answering iron.

NVMe nodes with engineered headroom, your database included, and protection that restores to the minute. Monitored 24/7, answered same-day.