Anyone can write the software. Fewer will still be running it in year three.
Custom business applications, integrations and the delivery platforms underneath them. Built by the same engineers who own your infrastructure, so there is nobody to hand the problem to.
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.
Tell us what the business actually needs to do. The technology choice comes after that.
What we are usually asked to build
Where business software actually goes wrong
Almost never the algorithm. These four account for most of the software we are asked to rescue.
The developer left
One person built it, nobody else understands it, and there is no document that would let anyone else take over. The business now depends on somebody who no longer works there.
It was never scoped
Development started from a feature list rather than the process the business actually follows. Six months later it does everything except the thing that mattered.
The code is not yours
Source held in someone else’s account, builds on a laptop, deployment by hand. You are licensing your own software back from the person who wrote it.
Nobody owns the data
No backup that has been restored, no schema documentation, no retention rule. The application can be rebuilt. Five years of records cannot.
Nexus. Everything classic ERP promised, without the complexity it demanded.
Nexus is Stark Technology’s Corporate Operation Management platform. We did not build it to sell a demo. We built it because we needed it, we run our own company on it, and it now runs a major regional call-centre operation around the clock.
People and workforce
The part most businesses run on spreadsheets, and the first thing that breaks when headcount grows.
- Human resources: records, contracts, documents and org structure
- Payroll, allowances, deductions and end-of-service calculation
- Leave and attendance, with approval routing that matches your policy
- Recruitment and applicant tracking, from vacancy to signed offer
- Workforce management: shift scheduling, rosters and live tracking
Designed for shift-based operations where hundreds of people change schedule every week.
Commercial operations
Sales, purchasing and spend in one place, so the numbers reconcile without a month-end scramble.
- Sales workflow from enquiry to invoice
- Purchases, supplier records and approval limits
- Expenses and petty cash with a documented approval chain
- Warehouse and inventory movements tied to the same records
- Finance and profit and loss built from live transactions, not re-keyed
One shared dataset. Nothing is re-entered between modules.
Built to be audited
If you cannot prove who changed what and when, the system is a liability during an audit.
- Every record change written to an immutable audit trail
- Role-based access, so people see the records their job requires
- Approval chains that reflect your delegation of authority
- Reporting that reconciles to the underlying transactions
- Evidence exports that stand up in an ISO 27001 or financial audit
This is the part that matters when a regulator or auditor asks.
Engineered for scale and uptime
A platform a call centre runs on has no maintenance window. It is either up or the business stops.
- Load-validated at 10,000+ concurrent users
- High availability by design, with no single point of failure
- Horizontal scaling across application and database tiers
- Monitored to the same standard as the infrastructure beneath it
- Backed by the same 24/7 support agreement as your other Stark services
The same team runs the platform and the infrastructure it sits on. One number to call.
Deployed where your data is allowed to live
Some organisations cannot put personnel or payroll data in another country. That is a design input, not an obstacle.
- Cloud, on-premises, or a hybrid split by data classification
- Deployable inside your own data centre, on your own hardware
- Integration with your existing identity provider
- Data residency kept inside Egypt where the law or the contract requires it
- Migration from existing systems, with reconciliation before cutover
Relevant now that Egypt’s Personal Data Protection Law is in force.
A documented delivery platform, not a laptop and good intentions
We build client software on the same self-hosted platform we deployed for an enterprise outsourcing group. Source, builds and artefacts stay inside infrastructure you control.
Source control and CI/CD
GitLab, self-hosted. Branching policy, merge review and pipelines that build the same way every time, on a server rather than someone’s machine.
Private container registry
Harbor, with image vulnerability scanning before anything is promoted. You know what is inside every image you run.
Static analysis and quality gates
SonarQube, with security gates that block a merge rather than file a report nobody reads.
Observability from day one
Prometheus and Grafana, so performance and errors are visible before a user reports them, not after.
Identity and access
FreeIPA for central identity, so joiners and leavers are handled once and access is provably removed.
Then we operate it
The same engagement covers monitoring, patching, backup and incident response afterwards. Building it and running it are not two suppliers.
Software running in places that cannot afford downtime
Real Stark work. Client identifiers are withheld under confidentiality. Sector and scope only.
Corporate Operation Management
Nexus running a major regional call-centre operation around the clock, and Stark’s own business alongside it. HR, payroll, workforce scheduling, sales, purchases, expenses, warehouse and finance in one audited system.
Secure DevOps and SDLC platform
A complete self-hosted software delivery platform built on open source, replacing commercial SaaS and keeping all source code and build artefacts inside the client’s own infrastructure. In production across 8 projects, documented as ISO 27001, PCI DSS and SOC evidence.
Platform engineering and DBA
Database administration, DevOps and system administration for a national e-payment platform, delivered as part of the full solution stack rather than as an isolated development contract.
References can be provided directly, on request and with the client’s agreement.
Three things we will tell you not to do
The questions we get asked most
Is Nexus a product we buy, or a project you build for us?
Both are possible. Nexus is a working product with defined modules, so most clients start from it and we configure and extend it to match their process. Where a requirement is genuinely outside it, we build that part and integrate it rather than forcing a fit.
Can Nexus run inside our own data centre?
Yes. Cloud, on-premises or hybrid. We deploy it on your hardware where data residency, a client contract or the Egyptian Personal Data Protection Law requires the records to stay in the country.
We already have an ERP. Why would we look at this?
Usually we do not replace a working ERP. Where Nexus earns its place is the operational layer most ERPs handle badly: shift scheduling, live workforce tracking, applicant tracking, and the approval chains that sit between departments. It integrates rather than competes.
Who owns the source code?
For bespoke development, you do, and it lives in a repository you control from the first commit. We will not hold a client’s code in our own account. Nexus itself is Stark’s product and is licensed to you.
What happens after go-live?
The same support agreement as the rest of your Stark services covers it: monitoring, patching, backup, incident response and an agreed response time. We do not treat go-live as the end of the engagement.
Do you take over software somebody else wrote?
Yes, and it is a large part of what we do. We start with an assessment: what it does, what it runs on, what is undocumented and what is unsupported. You get that in writing before you commit to anything further.
Describe the process. We will tell you whether it needs software.
Ten minutes is usually enough to say whether this is a build, a product you should buy, or a process problem that no software will fix.
Sunday to Thursday, 9am to 6pm · Egypt, and remote support worldwide
Our other IT services
Every service below is delivered by the same team, under the same agreement.