The title of this post is a question I genuinely haven't answered yet. I built a platform to help parents of autistic children find and understand resources. I have no reliable measurement of whether it helped anyone. That honest uncertainty is where this post starts.
Why I thought technology could help
The problem was obvious to me: the information parents need exists — medical resources, therapy options, community organizations, educational frameworks — but it's scattered across dozens of websites, written in clinical language, and impossible to evaluate without expertise the parents don't yet have.
A platform that organized this information clearly, in plain language, with some sense of what to look for first — that seemed like a solvable problem. I started building it.
The first version was organized for me, not for parents
I categorized resources the way I naturally thought about the topic: by type of intervention (behavioral, educational, medical), then by age group. This made sense to me.
When I showed the site to people who'd had no involvement in building it — including one parent who had direct experience navigating these resources for her child — the first thing she said was: "I wouldn't know where to start." That was the most useful feedback I received from this entire project.
She didn't think about resources by type. She thought about them by question: My child was just diagnosed — what do I do first? What does my child's school need to know? What can I do at home? My categories answered none of those questions directly.
The content problem
Technical accessibility — readable fonts, sufficient contrast, working links — was easy to get right. Communicative accessibility was much harder.
Clinical language is the default for most resources on autism. It's precise and defensible, but it's alienating to read when you're emotionally overwhelmed and trying to understand something about your own child. Rewriting content to be warm and clear, without being condescending or imprecise, took far longer than building the platform itself.
The hardest part of building something for people is that you have to understand how they think — not just what they need. These are not the same thing.
So — can AI actually help?
In theory, AI could help significantly: personalizing resource recommendations based on a child's specific profile, translating clinical descriptions into plain language, helping parents articulate what they're looking for. But all of these capabilities depend on first understanding what parents are actually asking — which requires the kind of deep user research I didn't do before building.
My honest answer to the question in the title: technology can help, but only if the people building it start by listening rather than building. I did it in the wrong order. That's the lesson I'd carry into the next version.