--- title: "Frequently Asked Questions – Help & Answers | MS WebX" description: "Find answers to common questions about our services, accounts, projects and how everything works — browse the full list below." canonical: https://mswebx.com/frequently-asked-questions language: en robots: index, follow og_type: website type: faq language_alternates: en: https://mswebx.com/frequently-asked-questions.md cs: https://mswebx.com/cs/casto-kladene-otazky.md de: https://mswebx.com/de/haeufig-gestellte-fragen.md es: https://mswebx.com/es/preguntas-frecuentes.md fr: https://mswebx.com/fr/questions-frequentes.md hu: https://mswebx.com/hu/gyakran-ismetelt-kerdesek.md it: https://mswebx.com/it/domande-frequenti.md nl: https://mswebx.com/nl/veelgestelde-vragen.md pl: https://mswebx.com/pl/najczesciej-zadawane-pytania.md pt: https://mswebx.com/pt/perguntas-frequentes.md ru: https://mswebx.com/ru/часто-задаваемые-вопросы.md tr: https://mswebx.com/tr/sikca-sorulan-sorular.md zh: https://mswebx.com/zh/常见问题.md ja: https://mswebx.com/ja/よくある質問.md ar: https://mswebx.com/ar/الأسئلة-الشائعة.md ko: https://mswebx.com/ko/자주-묻는-질문.md pt-br: https://mswebx.com/pt-br/perguntas-frequentes.md hi: https://mswebx.com/hi/अक्सर-पूछे-जाने-वाले-प्रश्न.md content_hash: sha256:a6e5f15e9c178e3f4b41d9e4c5f61ec7acc600648174d2c8e83068fe4bbb9e12 generated: 2026-08-11 --- > **Machine-readable Markdown rendering.** This is a Markdown representation of a web page, > served for AI agents and crawlers as a token-efficient alternative to the full HTML. > Full interactive HTML version: https://mswebx.com/frequently-asked-questions > Canonical URL: https://mswebx.com/frequently-asked-questions > Available languages: en, cs, de, es, fr, hu, it, nl, pl, pt, ru, tr, zh, ja, ar, ko, pt-br, hi > Change detection: `content_hash` is the SHA-256 of this document's Markdown body — everything > below this block, whitespace-trimmed, excluding the front-matter. Compare it to detect a real > content change; `generated` moves on every render and is not a change signal. # Frequently Asked Questions Find answers to common questions about our services, accounts, projects and how everything works — browse the full list below. ## Featured Questions What does MS WebX do? About MS WebX MS WebX is an IT company. We build websites and web platforms, native mobile apps for iOS and Android, and we integrate AI into them — together with the SEO work that makes what we build findable. We also run our own products on the same platform we build for clients, which means the foundation is proven in production before anyone else depends on it. How does a project with MS WebX start? Working With Us You describe the goal, not the technical solution. We ask questions until the scope is actually clear, then put it in writing before anything is built. If we think what you asked for is the wrong way to reach your goal, we say so before you pay for it — that conversation is cheaper than the rebuild it prevents. Do the website and the mobile app share the same backend? The Platform Yes, and more strictly than that phrase usually means. The API is defined once in PHP, and both the TypeScript client for the web and the Dart client for the app are generated from that definition. Neither is written by hand — the linter rejects a hand-written API type. So the app cannot quietly drift out of sync with the server: a mismatch breaks the build instead of breaking production. Is the 18-language support real, or machine translation on top? The Platform Real. Every page exists as its own row per language, with its own text and its own URL — this FAQ lives at /frequently-asked-questions in English and at /de/haeufig-gestellte-fragen in German, not behind a translation parameter. That means each language can be written the way people in that market actually search, and it can be corrected in one language without touching the other seventeen. ## Browse by Category [ ### About MS WebX Who we are, what we build, and how you can check it for yourself. 3 questions](/frequently-asked-questions/about-ms-webx "About MS WebX") [ ### Working With Us How an engagement starts, how a project runs, and what happens after launch. 5 questions](/frequently-asked-questions/working-with-us "Working With Us") [ ### The Platform One codebase, many brands, 18 languages, web and native from a single contract. 5 questions](/frequently-asked-questions/the-platform "The Platform") [ ### Security & Data Where data lives, how it travels, who can reach it, and what gets logged. 5 questions](/frequently-asked-questions/security-and-data "Security & Data") [ ### AI in Practice What AI integration actually means once it has to ship and keep working. 3 questions](/frequently-asked-questions/ai-in-practice "AI in Practice") ## About MS WebX What does MS WebX do? MS WebX is an IT company. We build websites and web platforms, native mobile apps for iOS and Android, and we integrate AI into them — together with the SEO work that makes what we build findable. We also run our own products on the same platform we build for clients, which means the foundation is proven in production before anyone else depends on it. Are you an agency or a product company? Both, on purpose. What we sell is already running in our own products first — the multilingual page system, the API layer, the admin tooling, the analytics. Client work does not start on an untested foundation, and when something in the core breaks we feel it before a customer does. How can I verify that MS WebX is a real, operating company? Three ways, all independent of anything we say here. The imprint page on this site carries the legal company details. Our own products are public and can be used without contacting us. And the contact page reaches the same people who write the code — not a call centre in front of them. ## Working With Us How does a project with MS WebX start? You describe the goal, not the technical solution. We ask questions until the scope is actually clear, then put it in writing before anything is built. If we think what you asked for is the wrong way to reach your goal, we say so before you pay for it — that conversation is cheaper than the rebuild it prevents. How does a project run day to day? In small, visible steps. Work goes to a staging environment you can open in your browser whenever you want, and it reaches the live site only when you say so. Nobody is shown a finished product for the first time at the end of the project — by then it is too late to change anything cheaply. What happens after launch? What we build is meant to be maintained: dependency updates, monitoring, and changes as your business changes. Whether we keep doing that or your own team takes over is agreed as part of the project, not assumed afterwards. Nothing is built in a way that only we can touch. Can you take over a system somebody else built? We read the existing code first and then say plainly whether continuing it or replacing it costs less over the life of the system. Both answers happen — an inherited codebase is often perfectly fine and just badly documented. We do not recommend a rebuild by reflex, because a rebuild is the most expensive thing anyone can sell you. What do you build on, and how do you keep quality up? PHP on the server, TypeScript in the browser, Flutter for iOS and Android, MariaDB for data, nginx and FrankenPHP for serving. Current, widely known technology on purpose, so anyone can maintain it after us. Quality is enforced by machines rather than good intentions: static analysis runs at its strictest level with no suppression list, alongside more than two hundred project-specific rules that reject code the moment it drifts. ## The Platform What do you mean by "the platform"? One codebase with a shared core and a thin per-brand layer on top. A brand gets its own design, its own content, its own database and its own domain, while security fixes, performance work and new core features arrive everywhere at once instead of being copied by hand into each site. Adding a brand is configuration and content, not a fork. Do the website and the mobile app share the same backend? Yes, and more strictly than that phrase usually means. The API is defined once in PHP, and both the TypeScript client for the web and the Dart client for the app are generated from that definition. Neither is written by hand — the linter rejects a hand-written API type. So the app cannot quietly drift out of sync with the server: a mismatch breaks the build instead of breaking production. Is the 18-language support real, or machine translation on top? Real. Every page exists as its own row per language, with its own text and its own URL — this FAQ lives at /frequently-asked-questions in English and at /de/haeufig-gestellte-fragen in German, not behind a translation parameter. That means each language can be written the way people in that market actually search, and it can be corrected in one language without touching the other seventeen. Can one platform serve several brands without them interfering? That is what it is built for. Each brand has its own database, its own domain and its own content, so one brand cannot read another's data and a content mistake in one cannot appear in another. What they share is the engine: the routing, the API layer, the admin tooling and the security work. Which products does MS WebX run itself? Cannabivo is live: a search engine for social clubs, where visitors find and compare clubs by location and clubs manage their own listing. Steel is a reference for material equivalents aimed at engineering and procurement; it is in development and not open to the public yet. Both run on the platform described in this section. ## Security & Data Who owns the data in a system you build for us? You do. We process it to run the service you hired us for and for nothing else. It is not pooled with another customer's data, not sold, and not used to advertise to your users. If the relationship ends, the data is yours to take with you. How is data protected while it travels? Connections run over HTTP/3, and the domain is on the browsers' preloaded HSTS list — so a browser refuses to talk to it unencrypted even on the very first visit, before any redirect could be intercepted. The key exchange is post-quantum: X25519 combined with ML-KEM-768, chosen so that traffic captured today still resists decryption once quantum computers are able to attack it. How are passwords, keys and API credentials stored? Never in source code — that is enforced by the build, not by discipline: code containing a hard-coded secret is rejected before it can be merged. Secrets live encrypted in a settings store and are decrypted at runtime only where they are actually needed, which also means rotating a key is one change in one place instead of a search across a codebase. What gets logged, and can you tell who changed something? Yes. Application events, errors and administrative actions are written to the database rather than to text files scattered over servers, so they can actually be searched. Administrative changes record who acted, what was changed and when. A question like "who changed this price last month" has an answer instead of a guess. How do you handle cookies and consent? Non-essential cookies are set only after the visitor agrees, the choice is recorded, and every site we build has a page where it can be changed or withdrawn later. Declining is one click, not a hunt through a settings menu — and a visitor who declines still gets a working site. ## AI in Practice What does "AI integration" actually mean in practice? A model doing one defined job inside your product, where it measurably beats the alternative — sorting incoming messages, drafting text a person then approves, making search understand what someone meant rather than what they typed. It is not a chat bubble glued onto a finished website. The useful question is never "can we add AI" but "which repeated task is currently costing you time". Where does AI already do real work in your own products? In content. Our multilingual material is produced and kept current with model assistance and then reviewed per language before it goes public — never published straight from a model. That is also why we are careful when we talk about AI: we know from our own work exactly where the output is strong and where it still needs a person. What happens to our data if an AI feature is involved? Before anything is built we name which provider processes which data and what is retained, in writing. If that answer is not acceptable to you, the feature is designed differently or not built at all. No data leaves your system in a way you have not agreed to — and "we did not know it was being sent" is not an outcome we are willing to produce.