Hi, I'm Isaac.

After graduating from UW-Parkside in 2022, I spent a few years as a senior full-stack engineer inside FinTech and benefits management companies. Those teams made a lot of high-stakes decisions — vendor selection, tool migrations, platform tradeoffs — and the process was always the same: someone built a spreadsheet, everyone had opinions, nobody agreed, and eventually the loudest voice in the room won.

When AI tools started becoming part of the workflow, I thought that would help. It didn't. Asking ChatGPT “should we use Jira or Linear?” returned two paragraphs of hedging that told me nothing I didn't already know. The problem wasn't access to information — it was that single-voice AI is structurally incapable of honest disagreement. It's trained to be helpful, which often means agreeable.

VerdictBox is my answer to that. A jury of 12 AI personas — each with a distinct perspective, priorities, and reasoning style — that actually argue with each other across three structured rounds before delivering a verdict. You see the debate. You see who dissented and why. You walk away with a decision you can explain.

It's just me right now. I'm building this in the open, shipping fast, and talking directly to anyone who uses it. If something doesn't work or you have an idea, email me — I actually read it.