Michael Yang

In my previous life, I was a quant trader trading options.

When I started experimenting with AI agents, I needed to switch models quickly across providers. A trader's odc to find the lowest price kept pushing me to figure out which provider was the cheapest.

That sent me down the rabbit hole of comparing inference costs. I realized cost is not just headline input/output token price. A huge part of my spend came from cache pricing, cache-hit efficiency, routing choices, and provider-specific behavior.

I ended up building a system to optimize all of that. Now I’m turning it into Auriko.ai