Artificial intelligence tools have made automation accessible to everyone: in a few hours you can build something that works. But between a test that works and a system a company relies on every day lies the same distance as between weekend DIY and a job done to professional standards.
Why automating seems easy today
Modern tools let you create an automation in a few hours, with no coding: you just describe what you want to achieve. It's an enormous step forward, but it creates a dangerous illusion — that the tool alone is enough to get a professional result. In reality, the tool is only the last link in a much longer chain.
What a demo doesn't show: which company data goes into the tool and where it ends up, what happens in unexpected cases, who checks the results, who updates everything when the tool changes version. These are exactly the points where improvised projects break.
The risks of improvising
An improvised automation works as long as everything goes as expected. The problems show up later: sensitive data handled carelessly, results never verified, processes that jam at the first out-of-standard case, no documentation. The real cost isn't the setup: it's the error that no one catches.
A few concrete examples we see in our audits: the quote sent to a customer with a miscalculated price, the message with the wrong recipient's name, the confidential document uploaded into the tool without data-protection settings switched on. For a Swiss SMB there's also a regulatory layer: the nFADP requires care in how customer and employee data is handled — including by automation tools.
What ICT experience and project management change
An automation project done well is not a single installation: it's process analysis, choosing the right tool, data protection, testing on real cases, training people and maintenance over time. These are the classic disciplines of ICT and project management, applied to new tools.
ICT experience brings the right questions about security, integration and data: how the automation connects to the tools the team already uses, which settings protect company information, what should happen when something goes wrong. Project management brings the method: measurable goals, clear priorities, defined timelines, assigned responsibilities and a training plan so the team actually uses what has been built.
That's how we work at Dedalix: every project starts with a free 60-minute audit to map your processes, continues with tailored installation and configuration, and ends with an operating manual, team training and support. Those who choose the ongoing plan also get updates and optimisation every month, because tools change fast and an automation left to itself ages within a few months.
Frequently asked questions
Can I build my company's automations myself?
For personal experiments, yes, and it's a great way to understand the potential. For the processes your team relies on every day you need data protection, testing on real-world cases and maintenance over time: without experience, the risk of invisible errors is high.
How do I recognise a reliable automation provider?
They ask to see your processes before proposing solutions, they talk about data and security before tools, they set measurable goals, and they include team training and post-delivery support. If the proposal comes before the questions, that's a warning sign.
How much does it cost to work with professionals?
Less than an error costs. Dedalix starts at CHF 290 per hour for hourly consulting; full automation of one business area (LAUNCH) costs CHF 6,900 one-off, currently on a launch promotion. The initial 60-minute audit is free and with no obligation.
Is my company's data safe with these tools?
It depends on how they're configured. The business versions of these tools offer settings that don't use customer data for training, but they must be enabled and verified. For law firms, fiduciaries and medical practices, compliance with the Swiss data protection law (nFADP) is part of the initial analysis.