Notes on AI, automation and training
Practical write-ups from real projects — no hype, no theory for the sake of it.
20 September 2026
Why Most AI Projects in Companies Fail (It's Not the Technology)
Most AI projects in companies don't fail because the models aren't good enough. They fail because the wrong problem got picked, nobody mapped the real process first, and no one followed up after launch. Here's what actually breaks these projects — and how to avoid it.
20 September 2026
What Is an AI Agent (Really), and Does Your Business Need One Yet
"AI agent" gets used for everything from a chatbot with a prompt to a system that reads your CRM and takes action on its own. Here is what actually separates the two, with concrete examples, and an honest checklist for when you need one and when you do not.
20 September 2026
RAG explained without jargon: how to make an AI use your company's data
RAG is what lets an AI answer with your company's real prices, documents and history instead of guessing from what it learned on the internet. No jargon, just what it does and when it actually pays off.
20 September 2026
n8n vs Make: How to Choose to Automate Processes in Your SME
n8n and Make are the two tools we get asked about most. The short answer: n8n wins when you need custom logic and control over your data; Make wins when you want to move fast with native integrations. The real decision, though, is about your process first.
20 September 2026
How to Choose an AI Automation Consultant: 7 Signs They Know What They're Doing
Look for a fixed project price (not hourly billing), a first call built on real research into your process, honesty about fit even if it means turning you down, hands-on tool fluency, and a track record of systems running unattended in production, not demos.
20 September 2026
How Much Does AI Automation Cost in Spain: Real Ranges
Prices for AI automation projects in Spain range from a few thousand euros for a single well-defined task to tens of thousands for broader integration projects. The single biggest factor is complexity, not hours worked — here's what actually moves the number.
20 September 2026
ChatGPT, Claude or Copilot: Which AI to Use for Each Task in Your Business
There is no single 'best' AI assistant. ChatGPT, Claude and Copilot each excel at different tasks: writing and quick answers, deep reasoning and long documents, or native Microsoft 365 integration. This guide breaks down which tool fits which job in a business context.
20 September 2026
AI Training for Teams: Why Most Fail (and How to Fix It)
Most AI training for teams fails because it teaches generic theory disconnected from the tools people actually use, then disappears the moment the session ends. The fix is training built around your real workflows, with real tools, and follow-up support after the session.
20 September 2026
AI Assistants on Your Website: What They Actually Do
A good AI assistant on your website answers real questions about your product, qualifies leads before they reach your inbox, and books calls without back-and-forth emails. It won't rewrite your funnel overnight — expect weeks of tuning, not instant conversions, before it earns its place.
20 September 2026
5 Processes to Automate with AI This Quarter
Lead routing, invoice reconciliation, client onboarding, recurring reporting and ticket triage: five processes most small businesses still run by hand, and what it actually takes to automate each one this quarter.