trending

Agents Observability for Product Teams

I built an internal tool to stop my product team from begging developers for data.

Friends saw it. They asked to pay for it. So next week I'm launching it on Product Hunt.
Here's the story:
Observability tools are built for devs. Dashboards full of latency graphs and stack traces.

Great if you're debugging. Useless if you're a PM trying to understand what users actually did in your product.
Our product and CX team kept asking engineering the same questions. Engineering kept answering with tools nobody outside engineering could read. So I built something simpler. Focused on the questions non-devs actually ask.
I shared it with a few friends running their own teams. Every single one asked to use it. Some offered to pay. Some asked to bring their own API key. That's usually the signal.
So I designed a logo, built a landing page, and gave it a name: Dimies.

Launching on Product Hunt next week.

Dimies - Observability tool for product teams

LLM observability today is built for engineers: tokens, latency, traces. Dimies translates raw AI-agent conversations into language product, CX, and leadership teams act on: Topics, resolution rates, out-of-scope demand, security events, and user frustration. It's the layer between eng tooling and business decisions: no dashboards to configure, no conversations to tag manually. Integrate in five minutes and know exactly where your agent wins, fails, and gets attacked.