Skip to main content
Currently taking on new projects. Free 15-minute call

Service

Web development, when a website isn't enough anymore

Logins, databases, workflows, interfaces. Software that runs in the browser and solves a concrete problem.

Book a free 15-minute call
Adidas, Haufe Group, Intalento, BMW Startup Garage, Groupon, Milon, adorsys, itsmydata, Mönius, shiftschool, siegfried und roi, PhysChem Analytics

In a nutshell

I build custom web applications and rebuild existing sites technically when they've become too slow or too rigid. The stack is Laravel and PHP on the backend, modern JavaScript on the frontend. The price depends on scope, mostly on the number of edge cases, and I quote it as fixed price, by effort or a monthly retainer, your choice, after the first call. I handle operations and ongoing development on request.

YouBookings arrive by email, end up in Excel, and someone types them into the invoicing tool. Errors included.
RafaelI build you a tool for exactly this workflow: the booking comes in, you approve with one click, the invoice goes out. No retyping, no errors. We start small with what you need daily, then expand.
Customer area with loginBookings and adminCalculator or configurator for your websiteInternal tools for your team
First version of a product ideaConnect your programsOverview and reportsApp for the phone

Not with me: Everything at once before anyone uses it. Software tied to a single vendor.

Past a certain point, a nicer website doesn't help anymore, only software does. That point is usually reached when workflows live in spreadsheets and email chains, when someone types data from one system into the next, or when an idea needs a login and a database just to work at all.

What I build

Custom web applications, meaning software that runs in the browser and models a concrete workflow. Booking and management tools, customer portals, calculators and configurators, internal tools nobody sells off the shelf. Plus MVPs for product ideas that first need to prove someone will use them.

And the other half: making existing systems faster and easier to maintain. A site that takes four seconds to load loses customers before they've seen anything. That's measurable and usually fixable in days, not months.

The proof is two products of my own

You can talk about references, but running software makes a better case. I built two applications entirely myself and operate them:

zeit is time tracking and billing for freelancers. Multiple clients, hourly rates, fixed prices and retainers, kanban for projects, expenses, export for the tax advisor, automatic bank transaction sync. Laravel at the core, plus a desktop app and an iOS connection. Hosted in Germany.

events plans parties and club events: RSVPs, potluck lists, carpooling, photos, all without guests having to register. Also Laravel, plus a mobile app built on the same codebase.

Both are not demos but live operations: with real users, real data, deployments, backups and everything that comes after. If you want to know how I work, sign up there and try it.

How I work

First the cut. The most expensive decision in a software project is packing too much into the first version. So I push hard on what really has to be in version one, and put the rest on a list for later. An application that runs and gets used is worth more than one that stays stuck in endless work.

After that: prototype before code, build in two-week increments, tests for the paths where a bug actually hurts. The source code belongs to you and sits in a repository you have access to from the start. The stack is deliberately standard, so you can hand it to someone else at any point.

What it costs

A number upfront would be a guess, so there isn't one. What determines the price isn't the number of screens but the number of edge cases: roles and permissions, payment processing, interfaces to other systems, and multiple languages are what multiply the effort.

In the first call you get a rough sense of whether your project sits at the lower or upper end, and whether it needs software at all. Sometimes the right answer is an existing tool plus two hours of setup. I'll say so if that's the case.

How it works

  1. Problem instead of feature list

    First the question of which workflow actually hurts today. Only then which software solves it. Often the answer is smaller than expected.

  2. Scope and cut

    What belongs in the first version, what can wait. An MVP that runs and gets used beats a full version that never ships.

  3. Prototype

    A clickable draft of the key workflows before any code exists. Changes are cheap here, expensive later.

  4. Build in increments

    Every two weeks something you can touch. No flying blind until final sign-off.

  5. Test and hardening

    Automated tests for the critical paths, checks for common security gaps, load testing when needed.

  6. Operations

    Deployment, monitoring, backups, updates. Hosted in Germany if you need that.

Checklist

Clear scope for version one, in writing
A visible milestone every two weeks
Automated tests for the critical workflows
Source code belongs to you, repository access from day one
No lock-in: the stack is standard, any developer can take over
Privacy by design, hosting in Germany possible

Frequently asked questions

What does a custom web application cost?

No honest answer exists without a call, and anyone who gives you one anyway is guessing. The price depends less on the number of screens than on the number of edge cases: roles, permissions, payments, interfaces. A tightly scoped internal tool and an MVP for real users are worlds apart because of this. After the first call you get a rough estimate, and after that a fixed price, by effort or a monthly retainer, your choice.

What's the difference to a normal website?

A website shows content. A web application processes something: users log in, enter data, data gets stored, calculated and returned. As soon as login, a database or permissions come up, it's an application, and the effort is in a different league.

What technologies do you work with?

Laravel and PHP on the backend, because it's fast, stable and easy to maintain for applications of this size. On the frontend, modern, native JavaScript, with Vite and Alpine when needed. A framework only gets added if the project truly needs it. All standard, so nobody depends on me.

Can you take over an existing application?

Yes, if the code is readable. I look at the project first and tell you honestly whether taking it over makes sense or whether a rebuild ends up cheaper. Both happen.

Do you also do pure performance optimization?

Yes. When an existing site is too slow, that's often its own small project: measure load times, find the causes, fix them. Details are in the Core Web Vitals guide.

Who operates the application after launch?

I do, on request. Deployment, updates, backups and monitoring as a monthly flat fee. Or your own team, in which case they get a handover and the documentation for it.

Related guides

Automation that gives you back hours every week

Workflows that sort emails, prepare quotes, and move data between your tools. No enterprise software, no AI …

Web design from one person who also builds

No handoff to subcontractors, no project manager in between. You explain it once, then the site gets built.

A brand that holds up in the browser too

Logo, colors, typography, application. Developed by someone who builds the website himself afterward and …

Landing pages that actually bring conversions

One page, one goal, a clear structure. I build landing pages that turn visitors into inquiries, buyers, or …

WordPress websites, without the page-builder compromises

Custom theme, a clean classic editor with ready-made sections, honest maintenance. Direct with a freelancer, …

Last updated: August 26, 2026