Notes from building real systems, shipped to real clients.
I'm Kritish. I run Kyrex and ship cloud and AI infrastructure for working businesses. GeekOnPeak is the long-form record of what I learn: case studies of production builds, and the posts that explain the moving parts. Everything here is free to read.
Distributed systems, worked through by hand.
A sequence on consensus and partition tolerance: what the papers claim, what the implementations actually do, and the places the two part company. First part lands in a couple of weeks.
Watch this space →One build not being named yet.
The one worth waiting for. It is on the bench now and the write-up follows about two weeks behind it — including the parts that did not work, which is most of what made it worth writing.
Watch this space →The newsletter is still in the books.
Not a publication yet, a list. It starts once there is enough here worth sending, and the first issues reach the list before general signup. Nothing is kept behind it.
Join the waiting list →