Spencer Pratt on Agentic RAGs + Chicago Reccs for Newbies
Description
Before he ever cried on the red line, Spencer Pratt broke his own RAG index.
In this episode of Detection Dispatch, Spencer Pratt (not The Hills one...this one writes detections, not drama) joins Dispatch to talk through what it really takes to operationalize agentic AI in the wild. From L1/2 triage to risk scoring, Spencer walks us through building a homegrown RAG system on top of Azure, complete with semantic search, vector embeddings, and even one risk score that always returns “zero” (because he told it to).
We get into:
– OpenAI in production for alert history correlation & analysis assist
– How to hallucination-proof your enrichment
– Why DNS exfil is still too weird for your LLM
– And why automation shouldn't make the decisions, but can help you decide faster
Also in this episode, you get a bonus:
🥲 Chicago starter pack of reccs for newly promoted SOC analysts
🍕 Bottomless brunch + skyline bike rides with the fam
🎮 Retro arcades and ramen bars that go harder than your SIEM
Detection Engineering Dispatch features candid conversations with security teams at top companies on how they build, measure, and scale world-class detection programs.
United States
