Most developers who build software for care organizations have never worked in one. Most people who understand care work can't build software. I do both, and that turns out to be the whole value.
I'm Valdrex, based in Los Angeles. I'm training as a marriage and family therapist at CSUN, and I've spent time working inside behavioral health and residential treatment settings, the kind of places where the work that matters happens alongside a surprising amount of paperwork, phone tag and manual coordination.
Somewhere along the way I started building software to fix the parts that frustrated me. That turned into Miwa, an AI-native EHR and practice assistant for therapists, which is to say I'm building the tool for the profession I'm training in. Alongside it sits Pwosesis, a privacy-first AI writing and thinking space for anyone working or studying in mental health.
Then other people started paying me to build theirs: an operations platform that carries home-health and hospice agencies from referral to paid claim, a per-diem marketplace for nurses and aides, and a placement referral service in Las Vegas. Same pattern every time. Someone who knows their work cold, buried in a process that software should have taken off their hands years ago.
I'm not going to tell you I'm a senior software engineer, because I'm not. What I am is someone who can sit in a room with clinical and administrative staff, understand what's actually happening in their day, and translate it into working software, using modern AI tooling to move fast, and bringing in specialist engineers when a project needs expertise I don't have.
The gap I keep finding is that care organizations get sold software built by people who have never done the work. It technically functions and nobody uses it, because it doesn't match how the day actually goes. I'd rather build the smaller thing that fits.
Los Angeles, California
Marriage & Family Therapy, CSUN
Behavioral health & residential treatment
Five live platforms, three for clients
Half the projects people ask me for shouldn't be built as described. Finding that out early is worth paying for.
A narrow tool in daily use beats an ambitious platform that never launches. I'd rather ship the first one.
I say no to work I can't deliver well. That costs me projects and saves clients from bad ones.
Documentation, training and a clear support arrangement, not a system only I understand.
I'm always up for a conversation about a process that isn't working, whether or not it turns into a project.