Pinned
in a year this is gonna feel obvious.
agents that quietly read their own production failures → turn them into better evals → and keep getting sharper without you manually babysitting every trace.
early version of that is already here and it’s kinda wild to see working.
00:00
01:34
Introducing Adaline 2.0 - The Agent Self-Improvement Layer
Adaline turns Traces into Behaviors,
Behaviors surface Issues,
Issues become auto-generated Evals + Data,
Adaline then generates new agent candidates and tests them.
You review the winners and ship!
