The Intelligence Is in the Architecture

I often compare this moment in AI to the early days of electricity. Electricity itself wasn’t the invention. It was a powerful force that humans had to learn how to harness, contain, and shape into systems that could actually serve us. The lightbulb mattered because someone understood how to contain that raw energy into something useful. I see AI in much the same way. The large language model is not the finished product. The real work is in designing the architecture around it and understanding how to contain this powerful force where human wisdom and clinical judgment must remain firmly in control.

Elah was built from the ground up around the work itself. A lot of people are surprised to discover that I coded Elah myself, of course, with the help of Claude, GPT and a few other LLMs that are brilliant for coding. Every line of code was carefully considered and methodically designed to emulate how we work.
Elah was not built around the novelty of AI.

Not around a cute bot. Not around the idea that adding a chatbot interface to a website suddenly creates intelligence.

It grew from countless experiments and lots of unanswered questions.

Questions like:

How do I gather better information without forcing people through exhausting forms or making them feel pressured?

How can I better prepare for sessions without losing presence during the pre-talk?

How do I keep important client information from becoming fragmented?

What can I do to ensure my clients information remains anonymous, private and secure?

I had so many questions to consider before I started writing a single line of code and by the time I completed the first iteration of Elah in early 2025, another huge question emerged:

How can Elah help practitioners refine their techniques by learning from their own work history?

That, right there, was an even deeper layer.

Over time, a practitioner’s own reflections may reveal patterns too. Perhaps certain approaches consistently produce stronger outcomes. Perhaps a practitioner repeatedly notes uncertainty in a particular type of case. Perhaps specific techniques appear more effective with certain client patterns. Perhaps there are areas where additional training could strengthen the practitioner’s work.

That is not just case management. That is the beginning of something much more interesting: technology that can help practitioners understand not only their clients, but their own evolving practice.

The Work Is Just Beginning

We are entering a period where thousands of AI products can be created quickly. Many will look impressive. Some will be useful. Others will be little more than familiar software with an AI label attached.

The question practitioners should ask is not simply:

“Does this use AI?”

The deeper questions are:
What problem was this actually built to solve?
Does it understand my workflow?
What happens to the information after it is collected?
Does context carry forward?
Does the system support my judgment or attempt to replace it?
Was privacy considered from the beginning?
Does this help me become more present with my client?
Does it help me understand what is changing over time?

Those questions matter because real professional intelligence is not created by connecting a language model to a chat box. It comes from understanding the work deeply enough to know where intelligence actually belongs.

That is what I have been building with Elah.
Not an AI hypnotherapist. Not a replacement for human wisdom. Not another instant “Create Your Own” chat bot in a box.

A system designed around the way hypnotherapists actually think, prepare, observe, reflect, personalize, and help people transform their lives. Because AI can be built in an afternoon. Understanding the lived experience and actually doing the work takes much longer.