Services · Cloud and infrastructure
Infrastructure you can watch, scale and recover.
For products and businesses that have outgrown a single shared host. We set up AWS, DigitalOcean or a plain VPS, with backups, monitoring and staging, and we run it afterwards.
- 2014
- Centralised cloud reporting since
- 125+
- Restaurants and cafés on Ethos
- 16
- Years building technology
Hardened infrastructure
ApplicationLIVE
ContainersRUNNING
ComputeMANAGED
Storage + backupRECOVER
Until it breaks
Hosting that only works until it does not
Shared boxes, forgotten backups and no one on call are how quiet nights end.
Review hostingHost
The application lives on one server nobody has logged into since the developer left.
Backup
Backups exist, in theory. Nobody has ever restored one.
Bill
The cloud bill grows every month and no one can say which service is costing what.
How it ships
Inventory, then harden
We list what runs today, then move it somewhere we can watch.
host/01
Understand
We inventory what runs where, who has access and what would break if the server disappeared this afternoon.
host/02
Design
The target set-up is drawn on one page with monthly costs beside each box, so you approve the bill before the build.
host/03
Build
The new environment is built alongside the old one, data is migrated with a rehearsal, and cut-over happens in a quiet window.
host/04
Stay
We keep the servers patched, the backups tested and the monitoring answered under a monthly arrangement.
What you can explain
Servers you can explain
Provision, monitor, back up and recover — without a mystery vendor in the middle.
Environments and deployment
Separate staging and live servers with SSH-key access and Git-based deploys, so a release is a repeatable step, not a copy over FTP.
Backups you have seen restored
Scheduled database and file backups kept off the server, with a restore drill before we call them done.
Monitoring and alerts
Uptime, disk, CPU and error-rate alerts going to a person, so you hear about the problem from us and not from a customer.
Database synchronisation and reporting layers
The pattern behind Ethos since 2014: outlets writing locally and syncing to one cloud address that holds every report.
Security basics done properly
Firewalls, TLS certificates that renew themselves, patched operating systems and access that is removed when people leave.
Primary deliverable
Ships with the engagement
01 · Lead outcome
The right platform, sized honestly
AWS when you need its services, DigitalOcean or a VPS when you do not; we size for today’s traffic and plan the step up.
- 01Environments and deployment
- 02Backups you have seen restored
- 03Monitoring and alerts
- 04Database synchronisation and reporting l…
In production
Work we can put a name to — or the method we will use until we can.
RestaurantFresspresso HospitalityFrom four outlets to 50+ cafés on one billing platform. ProductEthosBilling that never blinks.
Decide the shape
Cloud cost and ownership
Who holds the accounts, what a migrate costs, and what support covers.
DigitalOcean or a VPS is cheaper and simpler for most web applications; AWS earns its complexity when you need managed databases, queues, object storage or multiple regions. ATPL sets up both and chooses by what the application actually uses, not by brand.
ATPL quotes the set-up of the environment, backups and monitoring as a project, and the running of it monthly; the hosting bill itself is paid to the provider from your own account, so you are never locked in. The expected monthly bill is stated before building.
Yes. Ethos, ATPL’s restaurant platform, has used centralised cloud reporting with database synchronisation since January 2014, holding the reports for every terminal in 125+ restaurants and cafés at one address, and is served from ethospos.atplcloud.com.
Usually with minutes of downtime rather than hours: the new environment is built in parallel, data is synchronised, a rehearsal cut-over is run, and the real switch happens in a quiet window with the old server kept as a fallback.