AI-powered web applications

What we mean by an application

Not a brochure site with a form on it. Software with accounts, records, permissions and state: a directory with listings people manage themselves, a booking flow that has to hold up when two customers pick the same slot, a dashboard that has to stay correct when the data behind it is messy.

The AI part is a feature inside that, not the product. Search that understands a question rather than matching keywords. Listings that classify themselves. A moderation queue that sorts obvious cases so a person only sees the ambiguous ones. If AI is not the right tool for a screen, we do not put it there.

Decisions we make early, because they are expensive later

The data model, before any screen is designed. Almost every painful rebuild we have seen traces back to a schema that was convenient in week two and wrong by month six.

Arabic and English from the first line, with right-to-left layout treated as a first-class case rather than a mirrored afterthought. Retrofitting this is far more expensive than doing it at the start, and in this market it is not optional.

Authentication, roles and permissions before features. And where the thing will be hosted, backed up and restored — a plan you can execute, not a hope.

Built to be handed over

Everything runs under your own accounts: your hosting, your database, your domain, your third-party keys. At handover you get the code, the documentation and a walkthrough with whoever will maintain it.

We would rather be re-hired because the work was good than retained because you cannot leave. If a project ends with your team unable to run what we built, we did it wrong.

What we will tell you not to build

The feature list that assumes usage you have not proved. The mobile app that should have been a responsive site. The custom system that is a worse version of a product costing a fraction of the build. We have had these conversations, they have cost us work, and the clients who heard them are still with us.

What you get

Working software delivered in weekly increments rather than a reveal at the end, a fixed written scope agreed before we start, documentation and training at handover, and full ownership of everything built.

What it costs

First working version — KD 1,950. A scoped first phase: the core of the application running in production, delivered in weekly increments rather than one reveal at the end, with the scope written and agreed before we start, documentation and training at handover, and full ownership of everything built. Paid in advance, before work starts.

Anything beyond that first version is quoted once we have both seen it working, because that is the first honest moment to price the rest. A directory, a booking system and an internal dashboard are not the same size of job, and one figure on a web page covering all three would be a guess.

Running — from KD 195 a month once it is live: hosting, monitoring, backups, and the changes real use forces. A twelve-month agreement billed quarterly at KD 585 every three months, each quarter paid in advance.

This is the largest service we sell and the one we are most likely to talk you out of. If a spreadsheet or an off-the-shelf tool does the job, you will hear that on the call.

Describe what you are trying to build and we will tell you what it would take, what it would cost, and whether it is worth doing. If we are not the right fit, you will hear that on the call rather than in a proposal three weeks later.