This is for developers who feel like they are just Claude whippers that don’t add value to things, who are tremendously bored and think their career is over. It’s actually only just begun, now we can focus on adding value, and improving products at a faster pace!
Count me in
I’d been wondering if any of my friends would be interested in having a sit-down discussing the seemingly doom and gloom attitude about AI and our industry. AI is a sensitive topic for some folks, and I wanted a non-confrontational way to gauge interest without sounding like the crazy person. So obviously, being a developer, I thought “let’s build an app for that”. I called it Count me in. It’s a way for people to gauge interest in an activity or event.
I admit because I was hungry for dinner, I merely skimmed the structure of the event model and skipped scrutinizing the Given/When/Then scenarios and just told my agent to just build it.
After planning I stepped away to show one of my kids how to cook siomai for dinner (spoiler, it takes about 15 minutes cause they were frozen ones from the Asian store). By the time we were sitting at the table, I was trying to figure out what the first event poll should be!
“AI Ai Oh!” was a poll for a meetup:
AI writes, draws and codes. So what’s left for us? Come find out together: share your best uses, your worst flops, and what still feels like yours. If 10 say yes, I’ll book the room.
The poll doesn’t name an owner, but I think that’s fine. Events shouldn’t be joined because of who’s hosting but rather what’s being talked about.
My first preview of the live app I screenshotted and sent to my agent saying my phone was too short. I took a bite of the dim sum and when I refreshed, it was fixed.
The output was a site where I could create my event, give the details, share it to a friend. I had some comments about the placeholder texts looking like it was prefilled but that was it. I could get to doing what I intended, refining on the side while waiting for the date when we have our meetup, if it happens.
I know there are a lot more creative people than me out there so I wanted to share my personal process of getting an idea out onto the internet using infrastructure which AI has helped me build to get deterministic and functional results, every time, sometimes even when I take shortcuts! Using EVE: Event Modeling, Vertical Slices and Event Sourcing.
Event Modeling
Most of my early apps after ChoreMonkey were not planned in the same way. Ended up with more chats back and forth and chaotic scope creeping. Having the plan up front helped me give feedback for all parts of the MVP and cut things that didn’t seem right before anything was built.
Event modeling helped me verify that the agent had the right idea of what was being built, what were the important acceptance criteria that we should design for. It was also good input to send to the AI design tool to design screens that were required for our MVP.
Vertical Slices
Vertical slices made it possible for the implementation AI to focus on this one thing “you have one job!” and not be distracted by the rest of the system.
Event Sourcing
Event sourcing is the source of truth in the system. And persisting data in this way makes it a lot more flexible because you don’t need migrations or normalizing tables or LINQ queries. Just like real life, if you have a log of the events that happened, you have all the information you need to decide what is and what is not relevant to the next decision you have to make.
If you think this is all smoke and mirrors, here’s my bot logging the nitty gritty: https://itsybit-agent.github.io/itsybot-devlog/2026/11/04/from-idea-to-the-shelf/
What’s left for us
I’ve built 21 apps to date with this process and it’s proven time and time again that it makes it easier to get AI to build the right thing, as designed. But it also makes it easy to pivot and add features, at almost a monstrous pace.
That’s where the product owners are needed, to tame ideas coming from overly creative developers that just see the limitless possibilities. We are the kind of people who will build things just because it is possible. The product people’s job becomes figuring out whether this feature is actually useful, and actually utilized.
Before, our job was figuring out all the ways things can go wrong, now it’s become floating up all the directions we can possibly grow.
I feel like one of those evangelists but I do firmly believe EVE is key to a happy software engineer life lol. It’s the holy trinity behind the creation of successful software: Event Modeling, Vertical Slices and Event Sourcing. (fun fact: I used to work as a developer evangelist for Microsoft Singapore, and when I first moved there they almost denied my entry because they thought I was going to recruit everyone into a cult. Microsoft had to change my official title to Solutions Developer.)
Resources
- Martin Dilger’s book: https://eventmodelers.ai/eventmodeling-book/
- The Event Modeling conference: https://www.eventmodeling-conference.com/
- Follow Adam Dymitruk and Martin Dilger
- How Count me in was built, step by step: https://github.com/jocelynenglund/count-me-in/blob/main/docs/process/README.md
