Building a Therapy App as a Solo Developer-Therapist

By Justin DeGarbo, MA, AMFT · Founder & Clinical Director, Cathexis

I didn’t set out to build an app. I set out to solve a problem I kept watching happen in the therapy room.

A client would come in, and over the course of fifty minutes, something would shift. They’d notice the tightness in their chest for the first time. They’d connect a jaw clench to a specific memory. They’d feel their shoulders drop two inches and realize they’d been bracing all week. Real somatic awareness — the kind that changes how a person relates to their own nervous system.

Then they’d come back the following week and it would be gone. Not the memory of the session — they remembered it fine. But the felt sense, the body-level awareness, had dissolved back into the noise of daily life. By Tuesday, the insight was a thought instead of an experience. By Friday, it was something they’d meant to practice but didn’t.

This happened with nearly every client. Not because they weren’t trying. Because somatic awareness is a skill that degrades without practice, and there was nothing between sessions to keep it alive.

I looked for a tool. I looked hard. What I found were meditation apps dressed up in neuroscience language, mood trackers that reduced complex body states to a number between one and ten, and CBT workbooks with a body scan bolted on as an afterthought. None of them started with the body. None of them tracked somatic patterns over time. None of them were built on the neuroscience frameworks I was actually using in session — Barrett, Friston, Panksepp, Damasio, Solms.

The tool I needed didn’t exist. So I opened Xcode.

The learning curve nobody warns you about

I should be clear about what “taught myself to code” actually looked like. It looked like three months of not understanding what a variable was. It looked like spending an entire Saturday trying to make a button change color and failing. It looked like reading Apple’s SwiftUI documentation and feeling the same overwhelm my clients describe when they first try to notice what’s happening in their body — too much information, no clear entry point, everything connected to everything else.

I’m a therapist. My training is in clinical psychology, neuropsychoanalysis, and constructed emotion theory. Nothing in my education prepared me for dependency injection, state management, or why my app kept crashing every time someone rotated their phone.

But here’s what I had: the same stubbornness that got me through opioid recovery, through a master’s degree at Pepperdine while working full-time in residential treatment, and through building a pain recovery program from scratch at a Malibu treatment center. I know how to sit with something hard and not leave.

So I sat with Swift. For months. Then for a year. Then for two years.

Why I didn’t hire a developer

People ask me this constantly. The honest answer has two parts.

The first is money. I’m a therapist working in residential treatment. I don’t have venture capital. I don’t have a tech co-founder. The budget for Cathexis is whatever I can set aside from my clinical salary after rent in Los Angeles, which is not a sentence that inspires confidence in investors.

The second reason is more important: I don’t think a developer could have built this correctly without being a clinician. Cathexis isn’t a wellness app with a therapy skin. Every intervention script, every decision tree in the adaptive engine, every pattern the system surfaces — these are clinical decisions. They require understanding not just what constructed emotion theory says, but how it applies when a client’s chest tightness has been stable for three weeks and then suddenly migrates to their throat. That’s a clinical observation. It requires clinical judgment to handle well.

If I’d hired a developer and handed them a spec, I would have spent more time explaining the clinical reasoning behind each feature than it took to learn Swift. And the translation would have introduced errors — subtle ones, the kind where an intervention technically does what the spec says but misses the therapeutic intent behind it.

So I built it myself. Every line of code, every therapeutic script, every screen. It took longer. It was harder. But the clinical integrity is mine to guarantee, and that matters more than shipping fast.

What the app actually does

Cathexis starts every session with an interactive body map. You tap where you feel something — tension, heaviness, activation, numbness, warmth, whatever is present. You’re not rating your mood. You’re not selecting an emotion from a list. You’re identifying where something is happening in your body before your brain finishes constructing the story about what it means.

This is a direct application of Barrett’s constructed emotion theory: sensation precedes the label. If you can catch the body signal before it becomes “I’m anxious,” you have more agency over what happens next.

From the body map, Cathexis selects an adaptive somatic practice. Not from a playlist — from a clinically designed library that responds to what your nervous system is actually doing right now. The practices are grounded in Panksepp’s primary affect systems and structured around Ecker’s memory reconsolidation conditions. They’re short — three to five minutes — because the goal isn’t relaxation. It’s building the capacity to observe and understand your own somatic patterns.

Over days and weeks, the app connects individual check-ins into patterns. The chest tightness that spikes every Sunday night. The jaw tension that correlates with specific people. The heaviness that lifts after certain practices but persists after others. These are the kinds of observations that usually take months of therapy to surface — not because therapists aren’t skilled, but because they only see you for one hour out of 168.

That’s the gap Cathexis fills. Not therapy. The space between therapy sessions where somatic awareness either grows or dissolves.

The hardest part

The hardest part wasn’t learning Swift. It wasn’t the architecture decisions or the late nights debugging crash reports. It was the content.

Writing therapeutic programs — each a thirty-day sequence of interventions, psychoeducation, and adaptive practices — while maintaining philosophical consistency across every single script. Every line had to align with constructed emotion theory. Every practice had to respect the distinction between sensation and constructed emotion. Every piece of psychoeducation had to be accurate enough that a clinician would nod along, and accessible enough that someone who’s never heard of interoception would understand it.

I ran a full philosophy compliance audit across every intervention script in the app. I found places where the language had drifted — where a script implied the body “knows” something (it doesn’t; it provides data that the brain uses to construct predictions) or where an intervention treated emotions as things to be “released” rather than constructions to be observed and understood. Each one had to be caught and corrected, because the language shapes the therapeutic experience. If the script says your body is wise, the user learns to treat body signals as oracular truth. If the script says your body is providing data that your brain is interpreting, the user learns to observe — which is the actual skill we’re building.

That audit took weeks. It was the most important work I did on the app.

What’s next

Cathexis is live. The six launch programs cover the territories where I see the most clinical need: anxiety and hypervigilance, low mood and withdrawal, grief and loss, overwhelm and emotional flooding, chronic pain and the body, and trauma and past experiences. More programs are coming quarterly.

The provider dashboard — a web application that gives therapists structured data about what their clients’ nervous systems did between sessions — is projected for 2027. If you’re a clinician, you can join the waitlist at cathexis.health/providers. Your input directly shapes what gets built.

I’m still working full-time as a primary therapist in residential treatment. I’m still writing code after my shifts end. I still can’t make a button animate the way I want it to on the first try.

But my clients now have something between sessions that keeps the somatic awareness alive. That was the whole point. Everything else — the architecture, the algorithms, the App Store listing — is in service of that one clinical observation: the people who get better are the ones who learn to read their own bodies. Cathexis makes that skill accessible to people who can’t afford $200+ an hour or don’t have a therapist who speaks their body’s language.

That’s why I taught myself to code. That’s what Cathexis is for.

Cathexis is a somatic therapy app grounded in neuroscience. Start listening to your body today →