$1M to $100M run rate in four quarters. I have been staring at the graph for a while and it still does not feel real.
From the infrastructure side, it was less a growth story and more a capacity story. Agents needed real machines, thousands at a time, up in milliseconds, with state that survives. Every week the fleet had to be bigger and faster than the week before, and it was.
None of it happens without the team. The dedication over the past year, the odd hours, the refusal to let things break under customers. Big thanks to every one of you!
From the infrastructure side, it was less a growth story and more a capacity story. Agents needed real machines, thousands at a time, up in milliseconds, with state that survives. Every week the fleet had to be bigger and faster than the week before, and it was.
None of it happens without the team. The dedication over the past year, the odd hours, the refusal to let things break under customers. Big thanks to every one of you!
We just crossed $100M in run rate.View quoted post on X
Four quarters ago, we were at $1M. Then this year alone, we grew 37.5x.
Not long before that, we did a hard pivot- handed our old customers to competitors, rebuilt from scratch, and relaunched with a few months of runway left.
The bet was: models think on GPUs but work happens on CPUs.
So every agent that writes code, opens a browser, or runs an RL job needs a composable computer that's isolated, stateful, and ready in milliseconds.
$1M to $100M in about a year puts us in the same cohort as the likes of Anthropic and Cursor. It’s humbling to share that lane with companies we’ve learned so much from.