I have ADHD. I was diagnosed as an adult, which means I spent most of my career thinking I was just disorganized, easily distracted, and not as disciplined as I should be. Turns out I had a neurological explanation for all of it — not an excuse, just a context that helps you build better systems.
And one of the systems I needed most was a calendar that didn’t make my brain hurt.
Every calendar app I’ve tried — Google Calendar, Fantastical, Calendly, all of them — was built by people who don’t have ADHD. They’re beautiful. They’re functional. They’re completely wrong for how my brain works. Too much visual noise, no concept of time blindness, no way to handle the reality that an ADHD brain needs reminders in layers, not just one notification 15 minutes before an event.
So I built ADHDCal. And I built it using AI-assisted development. Here’s what I actually learned.
The Problem I Was Actually Solving
Before writing a single line of code, I spent a week just writing down every problem I have with existing calendar tools. Not “this feature would be cool” — actual, specific pain points:
- I’d see a meeting at 2pm and forget about it until 1:58pm because the notification came at 1:45 and I dismissed it reflexively
- I couldn’t tell the difference at a glance between a hard commitment and a “I should probably do this” item
- Time blocking never worked for me because the visual blocks looked the same whether the task was 20 minutes or 3 hours
- I needed to capture tasks mid-flow without losing my place — existing apps require too many taps or clicks for that
That list became the spec. Not a 40-page product requirements document — a list of specific problems I had, in my own words. It turned out to be the most valuable thing I did in the whole project.
Where AI Actually Helped
I built ADHDCal on Laravel (my home turf) with a Vue.js front end. The back end was straightforward — I know Laravel well enough that I didn’t need much AI help there. But the calendar UI is genuinely hard. Calendar rendering logic is notoriously gnarly, and Vue.js is not where I live.
This is where AI made the project possible. I described what I wanted in plain language — a custom calendar view that renders events with visual weight based on their type, with color-coded urgency rings instead of uniform blocks — and iterated on it with Cursor and Claude until it worked. What would have been two weeks of me fighting Vue documentation turned into three days of productive iteration.
The notification system was the other place where AI saved me significant time. Layered notifications (one hour out, 30 minutes, 10 minutes, 2 minutes) with different sounds for different priority levels sounds simple but the backend scheduling logic is tedious to get right. AI handled the boilerplate. I handled the architecture decisions.
Where AI Made Things Worse
This is the part people don’t talk about enough. AI-assisted development has failure modes.
Confident nonsense. AI will generate code that looks completely reasonable, compiles without errors, and does the wrong thing. It happened to me three times in this project. The worst was a date timezone handling issue — the generated code looked correct, handled DST, did all the right things, and still subtly shifted events by an hour in edge cases I didn’t catch until two weeks in. That cost me a day.
Context drift. If you’re working on a large codebase over many sessions, the AI loses track of decisions you made earlier. It will suggest things that contradict your existing architecture. You need to re-orient it constantly, which means you need to have a clear picture of your own architecture in your head.
Feature creep by suggestion. AI loves to suggest additions. “You might also want to add…” is a dangerous phrase when you have ADHD and a shiny-object problem. I had to consciously resist adding features the AI cheerfully scaffolded that I had no business building yet.
What the ADHD Angle Taught Me About Product
Building a tool for ADHD brains when you have an ADHD brain is both an advantage and a trap. The advantage: I know exactly what’s wrong with everything I build, because I use it every day. The trap: I kept second-guessing whether my own frustrations were universal or just my weird quirks.
What helped: I shared early builds with other people I know who have ADHD. Not a formal beta test — just “here, use this for a week, tell me what breaks.” The feedback was consistent in some places and wildly divergent in others. That told me a lot about what to prioritize.
The biggest insight: ADHD users are not looking for a simpler calendar. They need a more intentional one. Every design decision needs to be in service of reducing cognitive load — fewer choices, stronger visual hierarchy, faster capture. Simplicity isn’t the goal; intentionality is.
What I’d Do Differently
If I started ADHDCal over today:
- I’d invest more time upfront defining the data model. I changed the event/task schema twice mid-project, which is the kind of thing that costs you days when the codebase has grown. AI can help you sketch schemas quickly — use that early, not late.
- I’d write tests earlier. AI is excellent at generating test coverage. I left it too late and paid the price when refactoring notification logic.
- I’d keep a decision log. When AI asks “do you want to handle this with approach A or B?” and you pick A, write that down somewhere. Two months later you won’t remember why, and the AI definitely won’t.
The Real Takeaway
I shipped a SaaS application, solo, that addresses a real problem I have, in about six weeks of part-time work. Without AI tooling, I don’t think I’d have done it at all — not because I couldn’t write the code, but because the activation energy of starting would have been too high. Six weeks of spare-time project work was mentally manageable. Six months wasn’t.
That’s the real promise of AI-assisted development for solo operators and small shops: not that AI replaces expertise, but that it lowers the threshold for ambitious projects. Things that used to require a team are now possible for a person with a plan, solid fundamentals, and the right tools.
ADHDCal is still evolving. I use it every day. It makes my work life meaningfully better. And it exists because I stopped waiting until I had “enough time” to do it the traditional way.
If you have ADHD and want to talk about the app, or if you’re curious about what AI-assisted development could do for your next project, get in touch. I like conversations about real tools that solve real problems.