01App development
Building an AI app: what a demo doesn't tell you
AI in an app always demos well. The questions that decide whether it survives year two are about cost per user, offline behaviour, and consent.
Read the article →§ Index — Blog
Blog
Notes on the part of AI work that nobody demos: what it costs, where it breaks, and how to tell whether the technology is the right answer at all.
02Automation
AI automation or RPA: which one does your process need?
If your process fits in a handful of if-then rules, AI is the expensive answer. A practical test for telling the two apart.
Read the article →03Conversational AI
Building an AI chatbot that doesn't make things up
Most failed chatbots were never given a boundary, a source, or a way to say “I don't know”. Here is the order to build them in.
Read the article →04AI agents
What is an AI agent, and when do you need one?
A chatbot answers; an agent acts. That difference makes agents more useful and more dangerous — which is why boundaries matter more than autonomy.
Read the article →05Budget
What AI development actually costs
Three factors drive the price, and one of them is a running cost that scales with your success. What to budget for before you start.
Read the article →06Choosing a partner
How to choose an AI agency
Most AI projects fail on everything around the model, not the model itself. Six questions that tell you whether an agency has shipped AI into production before.
Read the article →§ Let’s Build
Got a harder question?
If your situation doesn't fit any of these articles, describe it. We'll tell you what we'd actually do.