BLOG
Web security, compliance and automation. Clear guides and real cases, written by the people who implement them.
The right order to design a bot that takes actions: what you should do, what people who get it wrong do, and why those mistakes are so common.
Read
2026-07-08
An AI bot that books, charges or cancels can wreak havoc if you build it wrong. The 6 design decisions that prevent 90% of the disasters, in plain terms.
Read
2026-06-25
Every supplier you onboard starts three regulatory clocks: GDPR, the AI Act and NIS2. Here's the real risk and how to keep it under control.
Read
2026-06-25
The EU AI Act also binds those who use AI through suppliers. Dates, risks and how to classify your systems before clients demand it.
Read
2026-06-25
Every supplier that handles your data needs a processing agreement. What Article 28 GDPR requires, the AEPD fines, and how to comply.
Read
2026-06-25
NIS2 requires managing your suppliers' cyber risk, with real verification. Status in Spain, who it applies to, and the fines.
Read
2026-06-25
Three regimes, one moment. How a compliance layer controls GDPR, AI Act and NIS2 right at supplier onboarding.
Read
2026-06-23
The worst kind of automation failure isn't the one that throws an error — it's the one that stays quiet. Why your automated process dies in silence, and how to find out the same day instead of three weeks late.
Read
2026-06-20
When your whole automation lives in one person's head, you have a business problem, not a technical one. What happens the day that person leaves, and how to avoid it.
Read
2026-06-17
When an automation does the same thing several times (charge, send, record), you have a duplicates problem. Why it happens and how to prevent it, especially with payments.
Read
2026-06-13
Your automations talk to outside services that change whenever they want. Why that breaks things and how to build connections that survive the hit.
Read
2026-06-10
Automating a messy process doesn't fix it: it makes it faster and harder to stop. Why you should tidy up first and automate second.
Read
2026-06-06
Every automation your business runs moves customer data through outside services. If you don't know where it goes, you have a GDPR problem. How to get it under control.
Read
2026-06-03
Many automations have a front door (a webhook) with no lock. An attacker can trigger it at will. What it is, why it's dangerous and how to close it.
Read
2026-05-30
AI can state false things with total confidence. They're called hallucinations. Why it happens, why it sounds so believable, and how not to trust it blindly in your business.
Read
2026-05-27
Someone hides malicious text inside something your AI will read, and your assistant obeys those hidden orders. What prompt injection is, with examples, and how to protect yourself.
Read
2026-05-23
An AI assistant can end up revealing customer data, internal pricing or its own instructions. Why it happens and how to stop your AI from being a leak.
Read
2026-05-19
AI tools are paid by usage, and that usage can spike without you noticing until the bill arrives. Why it happens and how to set limits before the shock.
Read
2026-05-15
Giving AI the power to decide alone on things that matter (charges, bans, firings) is tempting and dangerous. Where it can decide and where there must be a person.
Read
