We build the room the business runs in, and we build it to be maintained.
Servers, storage, virtualisation, racks, power and cooling, structured cabling and data centres, designed for the load you will have in three years, not the one you have today.
Engineering offices in Egypt, delivering on site across Egypt and Saudi Arabia, and supporting clients remotely across the Gulf, Africa, Europe and the United States.
Most people call. It is faster, and you speak to an engineer, not a form.
What a free infrastructure review looks at
Where infrastructure quietly goes wrong
Not dramatic failures. Slow, boring ones that are cheap to prevent and expensive to discover, and every one of them we have walked into.
Hardware past its support date
Servers running years out of warranty because they are still working. They are, until a disk or a board fails and the part has a six-week lead time, or none at all.
Power nobody has tested
A UPS with batteries older than the warranty, feeding a rack that has doubled in load since it was sized. The first real cut is the runtime test, and it fails.
A room that cannot breathe
Cooling sized for the original three servers, cabling that blocks the airflow, and a comms room that is really a converted store cupboard with the door kept shut.
Nothing written down
No rack elevation, no IP plan, no record of which licence covers what. Every change becomes an investigation, and the knowledge leaves when the engineer does.
What we design, build and hand over
From a single server refresh to a full data centre build. The scope changes; the discipline: survey, design, build, document, support. Does not.
Survey and design
Measuring what is actually there before drawing what should be. Skipping this is where budgets go wrong.
- Physical and logical survey of what you run today
- Capacity and growth modelled to a three-year horizon
- Power, cooling and floor-loading calculated, not assumed
- Resilience designed to the RTO the business agreed
- Bill of materials with alternatives at different price points
- Drawings and rack elevations issued before anything is ordered
You get the design whether or not you buy the hardware through us.
Servers, storage and virtualisation
The compute layer, sized honestly. Over-specification is as expensive a mistake as under-specification.
- Dell and HPE servers specified to the actual workload
- SAN, NAS and direct storage with growth headroom
- VMware or Hyper-V clusters with tested high availability
- Licence positions checked before purchase, not after
- Migration from the old estate with a rollback plan
- Performance baselined so future complaints have a reference
Backup and replication are designed alongside. See Backup & Data Protection.
Data centre, rack, power and cooling
The physical layer. We build these for clients, including at national scale, and we build them to be worked in.
- Data centre and comms room build, from empty room to live
- Racks, containment, PDUs and structured cabling to standard
- UPS and generator integration, with runtime actually tested
- Cooling sized to real heat load, with airflow managed
- Environmental monitoring, access control and fire detection
- Labelling and patching schedules that survive the handover
We build data centres for our clients; we do not resell space in our own.
Commissioning and documentation
The difference between a working installation and a supportable one is what gets written down.
- Structured testing and sign-off against the agreed design
- Rack elevations, IP plan, VLAN and cabling records issued
- Warranty, licence and support dates registered and diarised
- Runbooks for the routine tasks your team will do
- Handover session with your engineers, not just an email
- As-built drawings, not the design drawings renamed
The documentation is a deliverable with a due date, not a favour at the end.
Support, monitoring and lifecycle
Hardware ages on a predictable schedule. Planning for that is cheaper than reacting to it.
- Proactive monitoring of hosts, storage, power and environment
- Firmware and patch cycles scheduled into maintenance windows
- Warranty renewals and end-of-life flagged before they bite
- Spares strategy for the parts with long lead times
- Capacity reviewed against actual growth, not the original guess
- A refresh plan with dates and budget, agreed a year ahead
This can sit inside Managed IT as one agreement.
Infrastructure we have designed and built
Real Stark projects. Client identifiers are withheld under confidentiality. Sector and scope only.
National data centre
Designed and built end to end: racks, power, cooling, structured cabling, compute and storage. Under strict security and confidentiality requirements.
Platform behind a national e-payment service
Infrastructure supporting a live payment service, monitored around the clock, where a hardware failure is a public event rather than an internal one.
Server and storage refresh across sites
Aging estate replaced in phases with no unplanned downtime, virtualised and consolidated onto fewer, better-specified hosts.
Comms rooms and structured cabling
Cabling, containment and comms rooms built to standard in environments where dust, heat and vibration are part of the design brief.
References can be provided directly, on request, with the client’s agreement.
Three things we will tell you not to do
The questions we get asked most
Do you have your own data centre we can rent space in?
No, and we will not pretend otherwise. Stark builds and operates data centres for clients. Including at national scale, on their own sites or in facilities they choose. If colocation is what you need, we will help you specify it and manage the equipment in it.
Do we have to buy the hardware through you?
No. We are happy to produce the design and let you buy it however you prefer, including through an existing supplier. We would rather be judged on the design and the build than on the margin on a box.
Can you work with the equipment we already have?
Usually, and we will say honestly when something is not worth keeping. Plenty of estates need a smaller intervention than the client expects: more memory, a storage tier, or simply consolidation and documentation rather than replacement.
Can you do the work outside business hours?
Yes. Cutovers, migrations and power work are normally scheduled at night or at the weekend, and that is planned into the quotation rather than added later as a surprise.
Do you handle the cabling and the physical work too?
Yes. Structured cabling, containment, racks, PDUs, UPS integration and comms-room build are all in scope, done by our own engineers to a labelling and testing standard we hand over in writing.
Tell us what is holding your business up.
Ten minutes on the phone is usually enough to tell you what is near end of life, what is a genuine risk, and what can safely be left alone.
Sunday to Thursday, 9am to 6pm · Egypt, and remote worldwide
Our other IT services
Every service below is delivered by the same team, under the same agreement.