← Back to portfolio Case Study 03

Building a product with AI, without writing a line of code.

Irminsul  ·  Personal Project  ·  2024–Ongoing

Genshin Impact has around 60 million active players managing genuinely complicated rosters — dozens of characters, thousands of equipment combinations, team-based play that rewards deep optimization. The community has built tools to help, but they fall short in a specific way.

Existing tools are either spreadsheets that require real expertise to interpret, or generic tier lists that don't know anything about what a specific player actually owns. Neither one answers the question a player is actually asking: given what I have, what should I do?

I'd been watching this problem for a while. It was exactly the kind of gap I look for — a real unmet need with a clear audience and a tractable solution. I decided to build the tool myself.

I designed Irminsul — a companion tool that connects to a player's account via the game's API, reads their actual roster, and generates conflict-free team recommendations with plain-language explanations of the reasoning behind them.

The whole product was built without me writing a single line of code. I directed an AI through every decision: architecture, data sourcing, interface design, logic, testing. The AI wrote the code. I made every call about what the product should do and why — what problems it was solving, what the experience should feel like, where it could go wrong for users.

This is the same way I work with engineering teams. I define the problem space, make product decisions, specify behavior, and push back when something doesn't serve the user. The difference here is that the engineer was an AI — which meant I could move faster, iterate in real time, and maintain full ownership of the direction.

I'm not a developer and I didn't pretend to become one. What I did was use AI as a capable partner — the same way I'd work with an engineer — to take a clearly defined problem from concept to working product.

This project demonstrates a few things at once: that I can identify a real problem in a complex domain, that I can take a product from nothing to launched, and that I can use AI as a genuine force multiplier without losing the thread of what the experience should actually do for a person.

If you want to understand how I think about AI as a tool in design and strategy work, this is the clearest demonstration I can offer.

Live and in closed testing. Early feedback validates the core concept — users connect their accounts, generate recommendations, and navigate the interface without any guidance. Active development continues toward a broader launch.

Methods & Approaches
Product strategy AI-directed development Concept-to-launch Usability testing Iterative prototyping

Let's Talk

If you're working on something hard, I'd like to hear about it.

If you are at an organization where understanding people is treated as a real input to strategy, and not an afterthought, I would love to talk to you.

Based in Maplewood, NJ · Currently available for full-time or consulting roles at the Senior Director to VP level · New York metro or remote