Skip to main content

SERVICE LINE

Cloud & Data Centre

Tier-III-aligned facilities, colocation, hybrid cloud landing zones and disaster recovery you can actually invoke. Built and run from Kano by engineers who understand Nigerian power, Nigerian bandwidth and Nigerian data residency law.

Infrastructure designed for the conditions it will actually face

Engr. Bashir Lawal Gezawa leads the Infrastructure & Cloud practice, a team of forty-one engineers who design, build and operate the physical and virtual estate our clients run on. The practice grew out of a hard local truth: infrastructure designs imported wholesale from Europe or the Gulf fail in Nigeria, because they assume grid power that does not fail eleven times a month, ambient temperatures that do not touch forty-two degrees in April, and single-carrier fibre that does not get cut by a road contractor. Every facility we design starts from measured local conditions — grid availability, harmattan dust load, humidity, fuel logistics, carrier diversity — and works backwards to a resilience target the client can afford to sustain.

We build and refit data centres to Tier-III-aligned standards, meaning concurrently maintainable power and cooling: dual power paths, N+1 uninterruptible power supplies, generator sets with automatic transfer and at least seventy-two hours of fuel, precision in-row cooling with hot and cold aisle containment, VESDA aspirating smoke detection, clean-agent suppression, biometric access control, environmental monitoring and a building management system that tells you about a rising inlet temperature before it becomes an outage. Where a client has an ageing server room rather than a data centre, a refit is usually cheaper and faster than a new build, and we say so.

Alongside the physical estate we operate BrilliantCloud, our own virtualisation and colocation platform in Kano, and we design hybrid architectures that place workloads where they belong. Landing zones on AWS and Microsoft Azure are built to the vendors’ well-architected guidance with account separation, network segmentation, centralised logging, cost guardrails and infrastructure-as-code from day one. Data subject to Nigerian residency obligations stays in Nigeria; elastic, non-personal and burst workloads go to public cloud. Solar-hybrid power now supplies a meaningful share of our Kano facility load, which both reduces diesel cost and improves our clients’ carbon reporting.

Infrastructure services

From a concrete slab to a running landing zone, with the operational discipline to keep it running afterwards.

Data centre build and refit

Tier-III-aligned

Feasibility, electrical and mechanical design, containment, precision cooling, fire detection and suppression, access control, commissioning and integrated systems testing.

Colocation

Kano facility

Quarter, half and full racks with dual power feeds, carrier-diverse connectivity, smart-hands support and 24/7 escorted access for client engineers.

Hybrid and public cloud

AWS and Azure

Landing zone design, account and subscription structure, network segmentation, identity integration, cost guardrails and infrastructure-as-code in Terraform.

Virtualisation and private cloud

VMware and Proxmox

Compute and storage clusters, software-defined networking, self-service provisioning, capacity management and lifecycle patching of the hypervisor estate.

Backup and disaster recovery

Tested, not assumed

Immutable backup, offsite replication to a second Nigerian site, documented recovery runbooks and scheduled failover exercises with a written test report.

Power and cooling

Built for the grid we have

N+1 UPS, generator and automatic transfer switching, solar-hybrid arrays with lithium storage, in-row precision cooling and containment, and full environmental telemetry.

What we hold ourselves to

99.95%

Uptime attained

Hosted and managed estate, FY2025

15 min

Best-tier RPO

Continuous replication tier

4 hrs

Best-tier RTO

Warm standby with rehearsed failover

72 hrs

On-site fuel autonomy

Kano facility at full design load

Disaster recovery tiers

Recovery is a commercial decision before it is a technical one. These four tiers cover almost every requirement we are asked to meet. Indicative monthly costs assume a mid-sized estate of roughly twenty virtual machines and eight terabytes of protected data, and exclude VAT and carrier charges.

Tier Design RTO RPO Indicative monthly cost
Tier 1 — Backup and restore Nightly immutable backup with offsite copy; rebuild onto available capacity when needed 48 – 72 hours 24 hours ₦620,000
Tier 2 — Cold standby Reserved capacity at the second site, images pre-staged, configuration held as code 12 – 24 hours 4 hours ₦1,450,000
Tier 3 — Warm standby Continuously replicated virtual machines held powered-down; database log shipping every 15 minutes 4 hours 15 minutes ₦3,200,000
Tier 4 — Active-active Both sites live behind a global load balancer with synchronous database replication Under 15 minutes Near zero ₦7,800,000

How a migration runs

Whether the destination is our Kano facility, a public cloud landing zone or a hybrid of both, the sequence is the same. Nothing is moved until it has been moved successfully in a rehearsal.

1

Discover and inventory

Automated discovery across the estate produces an inventory of every server, application, database, interface and licence, with dependency mapping and measured utilisation over at least two weeks of normal operation.

2

Assess and choose a disposition

Each workload is assigned a disposition — retire, retain, rehost, replatform, refactor or replace — with the cost and risk of each option quantified. Roughly a fifth of workloads in a typical estate turn out to be retirable, which pays for a good deal of the migration.

3

Design the target and the landing zone

Network segmentation, identity integration, backup policy, monitoring, tagging and cost controls are all defined and built as code before the first workload moves, so that governance is inherited rather than retrofitted.

4

Pilot with a low-risk wave

A first wave of three to five non-critical workloads proves the tooling, the runbook, the cutover window and the rollback. Lessons from the pilot are written back into the runbook before wave two is scheduled.

5

Migrate in waves

Workloads move in dependency-aware waves, each with a rehearsed cutover, a data reconciliation step, a named business approver and a rollback point. Waves are scheduled around month-end, payroll and statutory reporting cycles.

6

Optimise and hand over

After stabilisation we right-size compute, tune storage tiers, remove orphaned resources and re-baseline cost. The client receives as-built documentation, runbooks, dashboards and, if wanted, a managed-service retainer.

Data residency and NDPR compliance

Personal data relating to Nigerian data subjects is subject to the Nigeria Data Protection Act and the NDPR framework administered by the Nigeria Data Protection Commission. Several categories of public-sector and financial data are additionally required by sector regulators to be held within Nigeria. Our Kano facility and our second Nigerian site exist so that clients can meet those obligations without giving up modern infrastructure.

Where a hybrid design sends any workload to a region outside Nigeria, we document exactly what data crosses the border, on what lawful basis, under which contractual safeguards, and how it can be repatriated. That analysis is delivered as a written data residency assessment before migration begins, not as an afterthought during an audit. Questions on residency and processing agreements go to [email protected].

Frequently asked questions

What does "Tier-III-aligned" mean, exactly?
It means the facility is designed to the concurrent maintainability principles of the Tier III standard — dual power paths, redundant cooling, and the ability to take any single component out for maintenance without shutting the load down — and is commissioned and tested against those principles. It does not mean the facility carries a formal Uptime Institute certificate unless we say so explicitly for a named site. We are deliberate about that distinction because the industry is not, and clients deserve to know what they are buying.
How do you handle grid failure and fuel supply?
Utility loss transfers to generator within ten seconds via automatic transfer switching, with the uninterruptible power supply carrying the load in between. The Kano facility holds seventy-two hours of diesel at full design load, with standing supply contracts from two independent vendors and a monthly load-bank test of the generators. A solar-hybrid array with lithium storage now offsets a meaningful share of daytime load, cutting both fuel consumption and generator run hours.
Do you test disaster recovery or just sell it?
We test it. Tier 2 and above include at least one scheduled failover exercise per year with a written test report showing actual achieved recovery time and recovery point against the contracted targets. Tier 4 clients typically test twice a year. If a test misses its targets we remediate at our cost and retest, because an untested recovery plan is a document, not a capability.
Can we keep some workloads on-premises?
Yes, and many clients should. Systems with heavy local integration, licence terms that penalise virtualisation, or regulatory constraints often belong exactly where they are. We design hybrid estates in which on-premises, colocated and public-cloud workloads share identity, monitoring, backup policy and network segmentation, so that the operating model is uniform even where the hosting is not.
What does colocation at the Kano facility cost?
Indicative pricing starts at ₦340,000 per month for a quarter rack with 2kW of dual-fed power, ₦610,000 for a half rack at 4kW, and ₦1,150,000 for a full rack at 8kW, each including redundant carrier-diverse connectivity, escorted access and remote-hands allowance. Higher power densities, additional cross-connects and dedicated cages are priced on application. Figures exclude VAT.
Are you tied to a single cloud vendor?
No. We are an AWS Advanced Consulting Partner and a Microsoft Solutions Partner for Infrastructure, and we build on both. Where the honest answer is that a workload should stay in Nigeria on our own platform, we say that too — it is frequently cheaper for steady-state workloads once egress and licensing are properly modelled. Our hosted platform is described at /solutions/managed-cloud, and sector examples appear at /industries/financial-services and /industries/health.

Server room at capacity, or a recovery plan nobody has ever tested?

We will survey the site, measure the real load and power profile, and give you a costed options paper covering refit, colocation and cloud — with the numbers that make the case either way.