Loading
A website and a web application are two different projects that get quoted as one thing far too often. The first has to load fast, read well in two languages and be editable by someone who is not a developer. The second has state, permissions and data, and belongs on an engineering plan. Novamind builds both, and separates them in the quote so you know which one you are buying.
Marketing and founder-level buyers who need a site their team can actually edit, and product owners who need a web application rather than a brochure.
The commonest complaint we hear about an existing site is not how it looks. It is that nobody can change it. The agency that built it has moved on, the content lives in a page builder whose export is unusable, and a price change means raising a ticket. A site you cannot edit is a site that goes stale, and a stale site is the one costing you enquiries.
The second pattern is a web application that was scoped as a website. Dashboards, logged-in areas, anything with permissions or workflow: those need a data model agreed before design starts, because retro-fitting one is the expensive way to discover it.
If you are in the Gulf and your customers read Arabic, that is a build decision rather than a translation task. Right-to-left layout affects navigation, forms, charts and every icon with a direction, and it is far cheaper to design for from the start than to add afterwards.
Four shapes, each scoped on its own. The first is where most projects start and the published AED 5,000 applies to it.
A fast, indexable site with a structured content model, so your team adds a page, a case study or a job posting without a developer. We model the content first, which is the difference between a CMS your marketer uses and a CMS your marketer avoids.
Both languages as first-class, with right-to-left layout, Arabic typography that is set rather than defaulted, and locale-aware numbers and dates. Content is modelled per language, so a page can exist in Arabic before its English twin without breaking the navigation.
Logged-in products with a real data model: roles and per-record permissions, server-rendered pages where they help the first load, and state that survives a refresh. Built on Next.js and Postgres, with the data model agreed in writing before any screen is designed.
When a browser is the right place for something that behaves like an app: installable, working offline against a local cache, syncing when the connection returns. Appropriate for field and warehouse use where a native app would be overkill and a plain site would be useless.
The stack is chosen so that a UAE development team can take the project on without us, which is the only durable protection against being locked in.
TypeScript, Next.js, React and Tailwind on the front, Postgres behind it. These are hireable skills in Dubai, Riyadh and Kuwait City, which means the code is maintainable by somebody other than the people who wrote it.
Every project ships with a CI pipeline, a staging environment you can open, and a production deployment. The staging URL goes to you at the start rather than at the end, so you are watching the build rather than receiving it.
We do not build on proprietary page builders or low-code platforms. They are faster for the first two weeks and a trap at the point you want to leave, because the export is not a codebase.
We build for both, and we are not going to promise you a number or a ranking.
The technical groundwork is standard on every build: server rendering or static generation where it suits the page, images sized and served in modern formats, a sitemap and robots file that match the routes that actually exist, canonical tags, and structured data where a page has a type schema.org understands.
What we will not do is promise a Lighthouse score or a search position. Scores move with the network the test runs on, and rankings depend on your competitors and on Google. Anyone quoting you a guaranteed position is selling you something they cannot deliver.
What we can do is measure. If you want search performance tracked after launch rather than assumed, that is a retainer conversation and we will say what we are measuring before we start.
Stated plainly, so nothing is a surprise at handover.
Guaranteed search rankings or a guaranteed Lighthouse score. We build for both and promise neither.
Ongoing content writing, photography or video as part of the build price. Those are separate and we will quote them separately.
Hosting and maintenance as an included extra. Deployment is part of the build; keeping it running month to month is a retainer.
On-site work. Discovery, design, build and launch all happen remotely over calls, a shared repository and a staging URL.
A published uptime guarantee. There is no SLA, so we are not implying one.
This site, the three product sites we run, and a published client case study are all things you can open and judge for yourself.
The GSG Academy platform is published as a case study: an indexable web platform at gsgacademy.co where every course, event and programme has its own page that deep-links into the mobile app, plus the admin console the academy team runs enrolment, payments and campaigns from.
Our own product sites, axionerp.co and ignisapp.co, are built on the same stack this service uses, and both carry the kind of thing we are describing here: bilingual content, structured data, calculators and tools that work without a page reload.
Judging a web developer on their own sites is fair, so we would rather point you at ours than at a gallery of screenshots.
The questions we hear before a web development project starts.