Automation, Workflows & AI
Maybe AI is hype. We’re doing it right anyway.
For us, automation and AI belong together. One moves information to where it’s needed, the other does something smart with it.
The potential is real, and it often hides in inconspicuous corners: data copied from one tool to the next. Reports compiled from scratch every week. Requests that wander through three inboxes before anyone answers them.
That’s exactly where we come in. We connect the tools and apps you work with every day, automate recurring processes, and build AI in where it makes a real difference: as an assistant working with your knowledge, as a feature in your product, or as a smart step in the middle of an automation.
Cleanly built and maintainable, so the solution keeps running for the long term.
What we offer
Workflow & Process Automation
We connect your tools with each other: your CRM with your accounting, HubSpot or Salesforce with your newsletter, your contact form with the right inbox. Built with platforms like n8n or developed as a custom solution. What’s manual work today runs on its own afterwards: transparent and adjustable at any time.
AI Integrations & Assistants
An assistant that pre-sorts incoming requests. A search that understands what you mean. We integrate AI features directly into your website, your application, or your internal tools. Embedded in what you already use, instead of yet another tool nobody opens.
AI Strategy & Use Case Discovery
Where is AI really worth it for you? We look at your processes, find the use cases with the best ratio of effort to benefit, and tell you just as clearly where it isn’t worth it (yet).
Feasibility, Data & Governance
Before anything gets built, we clarify the fundamentals: Is the necessary data available and usable? What is a model allowed to see, and what isn’t? Which providers meet your requirements for data protection and location? That way, your solution stands on solid ground.
Got a process in mind that eats up too much manual work? Tell us about it.
First structure. Then automation. Then AI.
Why this order? Because most AI projects don’t fail because of the model, but because of scattered knowledge and unclear processes.
That’s the principle we follow with our clients, just as we do ourselves. It’s not a rigid sequence: the three building blocks interlock, and they often take shape in the same project at the same time. But the foundation determines how far you’ll get.
Structure as the foundation
An AI assistant is only as good as the information it gets. That’s why we look at your processes and data first: Where does the knowledge live? Where is it maintained twice, where does it get lost? Sometimes the most valuable first step isn’t a new tool, but putting your house in order.
Automation as the connection
Automation moves information to where it’s needed, all by itself. No copying by hand, no updating three systems. We build it integrated into your existing environment instead of as an isolated solution, and without unnecessary dependence on a single provider.
AI where it makes the difference
On a solid foundation, AI works with your knowledge instead of guesswork. Automation and AI often live in the same workflow: a model summarizes, translates, or reviews, and the automation delivers the result to where you work. And because solutions like these are software, we stay on it: monitoring, maintenance, further development. For years, if you like.
→ More on this under Maintenance, Operations & Further Development
A selection of our clients.
Does this sound familiar?
- Your team regularly copies data by hand between systems that really should understand each other.
- You’d like to use AI, but nobody can tell you where to start and what it would cost.
- A first automation already exists, but it’s undocumented, and nobody dares to touch it.
- Management wants an “AI strategy”, and you’re looking for someone who gives an honest assessment instead of a sales pitch.
In all four cases: get in touch. The first conversation costs nothing but half an hour.
Which AI model is the best?
Articles about benchmarks appear regularly, showing that the latest model from Anthropic is the best, or that OpenAI has caught up or pulled ahead again. There are also sources dedicated to specific use cases that name the right models for them, such as SWE-bench, a benchmark for software engineering capabilities.
The problem: the article you find is sometimes a few months old, and so is the information in it. In the meantime, new models have arrived, the ranking has changed, and opinions differ. Or someone comments on LinkedIn: “What, you’re still using Claude? OpenAI has long since pulled ahead.” And depending on the current news cycle, the other way around too.
Put simply: there is no best model. But there might be a best model for a specific use case, at a specific price.
To be honest: We don’t have all the answers either, and the landscape is changing at breakneck speed. But we test continuously and are happy to help you with our experience.
There are different ways to try out different models. You can build your own tools for it, or take out a subscription everywhere. But there are also various providers and options that let you test different models side by side, billed purely by usage, meaning token consumption.
There’s also a provider from Switzerland that bundles the big and latest models and makes them accessible in a single subscription. If you’d like to consolidate your various AI subscriptions: have a look at Mono. This isn’t paid advertising, but at this point we should also mention that we know and appreciate the founder personally.
Our values
Whether it’s design, concept, interfaces, engineering, web development, WordPress, or indeed AI: our standard for quality stays high.
A deliberate approach to technology
We adopt new technology thoughtfully, not because of hype. We check whether something really holds up before we build on it.
Development & engineering in-house
Design and development come from one team, in-house. Nothing is outsourced.
Remote & distributed since 2013
We know digital work from more than a decade of our own practice, including the processes and tools behind it.
Honest advice as equals
We recommend what a project really needs, and tell you just as clearly what it doesn’t.
Do you recognize yourselves in some of these situations?
In a first conversation, we’ll look together at where automation or AI gives you the most leverage and whether your structure is ready for it. No strings attached and no buzzwords.
How we work ourselves
We don’t sell anything we don’t use ourselves
We’ve been working remotely and distributed since 2013, as a deliberately small team with big projects. Our system makes that possible: our knowledge lives in one place, and various automations keep it up to date in the background, built with tools like n8n.
- A meeting in the calendar automatically becomes a prepared entry in the project hub
- A client email becomes a task on the right project
- A code change on GitHub is documented before anyone asks about it.
Nobody copies anything by hand. And on this foundation, an AI assistant answers our questions from our own knowledge, even when the person who’d know best is on vacation.
AI tools are part of our everyday development work too: they support us in writing, reviewing, and documenting code and make us faster, without us handing over responsibility. The model landscape changes almost monthly. We test continuously to see what has actually improved, and that keeps us close to the topic.
On top of that, we build our own tools that use AI. For translations, we developed a two-step method: one model translates, a second refines the text. Which models do this best is something we test continuously. In our WordPress projects, AI features help create content and check its quality with publishing checks before anything goes live. And our own project dashboard contains an MCP server, an interface through which AI models talk directly to our tools: that’s how we ask in chat how our projects are doing and where maintenance is due next.
→ How we support projects over the years: Maintenance, Operations & Further Development
That personal side projects emerge along the way, like a hiking app that knows the weather for every waypoint when you’re planning a tour, or an app that recommends new content and even vacation destinations based on a favorite movie, is almost a given here. If you love building, you keep building after hours.
WordPress itself is also expanding its AI features right now. We’re following this up close and bringing it into projects where it fits.