Hi everyone, I'm building a memory layer for AI Agents to self improve. Think of it as a feedback loop where the production data gets fed back automatically so the resolution rate gets better over time, just like how a human learns on the job. Would love to chat if you find your company spending a lot of FDE time maintaining agents after deployment / escalation rate stagnates. We are very early so want to get as much feedback as we can! We just launched our website yesterday (https://trycognell.com) Here's my calendly if anyone wants to book a call directly https://calendly.com/jaylim95/30min