# Systems that shipped, and what they had to survive.

> Four systems in different industries — retail commerce, rental commerce, accommodation SaaS and industrial data — chosen because they show the same approach applied to problems that look nothing alike from the outside.

- **Email:** hello@zhajmohamad.com
- **URL:** https://zhajmohamad.com/projects/

## What the four have in common.

Different industries, same shape of problem: a business rule that the software has to hold, under load, without a person watching it.

- **The rule lives in the system** — Availability, pricing and stock rules are enforced by the platform rather than by a person who remembers them.
- **Integrations are queued, not hopeful** — Every cross-system message can be retried, traced and reconciled, because the other side will eventually be down.
- **Configuration beats deployment** — Onboarding a property, a clinic or a price tier is a record in the system, not a release from an engineer.
- **One engineer accountable** — Each of these was architected and built by the same person who then had to keep it running.

## What each one was called in for.

None of these started as a greenfield brief. Each started with something that had begun to fail.

- **A backend that could not take campaign day** — Frankenspalter — thousands of orders a day against catalogue indexing and an ERP that also wanted the database.
- **A rental sold like a product** — Dcey — a storefront where the date range had to become part of the transaction, not a note on it.
- **A platform that had to serve many operators** — Hujuzatk — properties, rooms and availability per tenant, with rules in configuration.
- **Production data nobody could see** — DORTEK — PLC signals that stopped at the panel until there was a gateway and a cloud to send them to.

## The case studies.

- **Frankenspalter (frankenspalter.ch)** — A Magento backend for thousands of orders a day, with ERP integration that runs while the store keeps selling.
- **Dcey (dcey.com.tr)** — Headless rental commerce: Magento GraphQL under a Next.js storefront built around date ranges.
- **Hujuzatk (hujuzatk.com)** — Multi-tenant property booking and management SaaS, where onboarding an operator is a record rather than a deployment.
- **DORTEK (dortek.com.tr)** — PLC lines through a gateway into cloud dashboards, feeding the stock and planning decisions behind them.

## Questions people ask.

**Why are there only four case studies?**
Because these four are the ones that can be written about openly and that cover genuinely different problems. Other systems — clinic scheduling, education platforms, watch retail — appear on the main site, and the rest sit under agreements that keep client names unmentioned.

**Can you talk about a system like ours before we sign anything?**
Yes. The first call is about your system, its constraints and the date — not a pitch. You get a straight answer on fit within a day, and a name to call if it is not me.

## Related specialisms.

- [Custom e-commerce development](https://zhajmohamad.com/ecommerce)
- [Magento / Adobe Commerce architect](https://zhajmohamad.com/magento-adobe-commerce)
- [Booking & reservation platforms](https://zhajmohamad.com/booking-platforms)
- [Property management systems (PMS)](https://zhajmohamad.com/property-management-systems)
- [Industrial IoT & PLC integration](https://zhajmohamad.com/industrial-iot)

---

Tell me what you are trying to build, automate or scale. You get a straight answer on fit within a day — and a name to call if it is not me.