18 Jul 2026
Why we stopped quoting cold start as an average
An average cold start is a number that describes nobody. Ours was 61 ms and the distribution was bimodal.
Tag: <span>postmortem</span>
Tag: postmortem
30 Jun 2026
A proxy buffered the first 4 KB and nobody noticed for six weeks
Streaming looked fine in aggregate because the total time was unchanged. Only the first token moved, and we were not graphing it per region.
8 May 2026
Mixture-of-experts pricing was wrong for four months
We were counting tokens against total parameters rather than active experts. Everyone affected has been credited; nobody had to ask.