The free 15-minute intro call checks whether we're a fit; the paid 90-minute KI-Sprechstunde then gives you a concrete recommendation for your first step. In the intro call we listen briefly: your role, your company, where it hurts. In the working session we go deep, test one concrete question for value and feasibility, and you leave with three clear next steps. No sales pressure, online, no slide decks.

The first step isn't a tool, it's a clear question: which process costs you the most time today? That's exactly where we start. In the free intro call we find the most valuable starting point together, instead of trying tools at random. Strategy first, tools later, which saves you weeks of detours.

For small companies AI often pays off fastest, because even one automated routine gives back noticeable time. You don't need a big budget or an IT department. We find the one process eating the most hours today (invoices, quotes, email triage) and make it leaner. Four hours back per week is a real difference for a small team.

We mainly work with Swiss SMEs in accounting, healthcare, construction, retail and church organisations, wherever there's a lot of admin. The common thread isn't the industry, it's recurring office work: documents, requests, scheduling, data. If your team spends a lot of time on those routines, our approach fits, regardless of sector.

No, you need neither coding skills nor your own IT department. We speak plain talk instead of jargon and handle the technical side in the background. Your job is to know your business, ours is to make the technology fit it and work day to day.

Good, then you're already 90% of the way there. An audit or workshop with us handles the last 10%: a clear plan on which tools stay, which go, and how your team actually uses them day to day. You're not starting from zero, you're turning the experiment into a system.

Yes, those are among our most frequent clients. An accounting office automates document processing, a parish its correspondence and admin, a trades business its quotes and scheduling. The tools are similar, the concrete benefit differs per business, and that is exactly what we find together.

A typical project runs in four steps: free intro call, KI-Sprechstunde or audit for the plan, building the first use case, then go-live with your team. We start small and visible: one process that matters, not a mega-project. You see a result early, then decide on the next step. That keeps the risk small and the progress tangible.

A first, simple automation is often live in one to three weeks, not months. It depends on the process: an email routing or appointment confirmation goes fast, a connection to your accounting takes a bit more. We deliberately pick a first step that shows impact quickly, so your team feels the benefit early.

After the intro call you get a clear recommendation: either the first use case directly, or an audit if there are several areas to tackle. We tell you honestly what's worth doing first and what can wait. You decide, we build. No contract for things you don't need yet.

Start with rule-based, frequent, high-volume tasks, such as invoice reminders, email triage or appointment confirmations. These tasks are clearly defined, error-prone by hand and add up to a lot of time, the ideal first win. Complex, judgement-heavy tasks come later, once the basics are in place.

n8n is an open automation tool for connecting software and automating workflows, and it can be self-hosted in Switzerland in a privacy-friendly way. Unlike Zapier or Make, we can run n8n on a Swiss server so your data stays in the country. You are also not tied to one vendor and do not pay per-action prices that explode with volume.

ChatGPT is a general assistant you operate yourself; a custom solution runs in the background of your workflows and knows your data and rules. With ChatGPT you ask and copy the result by hand. A custom solution fetches the information itself, does the step automatically and follows your rules. ChatGPT is the entry point, the custom solution saves time continuously.

Yes, we connect AI workflows to common Swiss business software like Bexio, SAP and Microsoft 365. Through open interfaces we pull data where it already lives and write results back, without you having to switch systems. If a tool has no interface, we usually still find a clean path.

We're not tied to one model and choose between ChatGPT, Claude, Gemini or a Swiss-hosted model depending on the task. For sensitive data we lean on Swiss or European options, for general tasks on whatever works best and cheapest. You get the right solution, not one vendor's favourite software.

You're not locked in: what we build is yours, including access, documentation and the know-how to run it without us. We use open tools like n8n that you can operate on your own, not a black box only we can maintain. If you want to take it all in-house, we hand over the complete setup. Plenty of companies stay in sparring afterwards because they want the thinking partner, not because they have to.

The free standard version of ChatGPT is not automatically compliant with Swiss data-protection law for sensitive business data, because your inputs are processed on US servers. For general, anonymous questions that is usually fine. The moment customer data, HR data or trade secrets are involved, you need a clean setup: ChatGPT Enterprise or the API with a data-processing agreement and training switched off, or a model hosted in Switzerland. We work out which path fits you in the KI-Sprechstunde, without legal jargon.

When your requirements call for it, we host your solution in Switzerland so your data doesn't leave the country. Our automations run on Swiss servers (Infomaniak), and for AI models we choose Swiss or European providers when data is sensitive. Where your data sits is a deliberate decision, not an afterthought.

For legally compliant AI use you mainly need three things: clear internal rules, clean handling of personal data, and transparency about which tools are allowed for what. You don't have to piece this together yourself. In a governance step we define with you which tools are allowed, how to handle sensitive data and who decides in case of doubt. Clear rules instead of fear of data protection.

Microsoft Copilot can be used in a data-protection-compliant way by Swiss companies when it is configured correctly, for example via the business version with the right privacy settings. What matters is the licence and setup: in the free or personal version you have less control, in the Microsoft 365 business version considerably more. We check with you whether your Copilot is set up cleanly before sensitive data flows into it.

Yes, after go-live we don't leave you alone, and on request we keep supporting you through ongoing sparring. You can run the solution yourself or keep us as a thinking partner who weighs in, sets priorities and helps with the next step. Six to twelve months of sparring is common, cancellable monthly.

We train your team in a hands-on workshop built around your real tasks, not abstract slides. In half a day or a full day we pick the right tools together, practise the first prompts and build a plan for next week. Your team leaves with more clarity than ten demos deliver.

Yes, we build it so your team can operate the solution day to day without calling us for every little thing. You get a clear guide and a short introduction so daily work runs without us. For bigger changes or new ideas we're here, but the day-to-day is in your hands.