DeepTempo blog

LogLM Trumps Frontier and Open-Source Models in SOCBench Cybersecurity Detections

Where are the Cyber SREs? Lessons learned from DevOps that can apply to today's necessary transformation of cyber security

The 36 Percent False Positive Problem with LLMs: Findings from Our Open Benchmark, SOCBench

The AI Attacker SNAFU - and how to spin a faster OODA loop to respond and win

Introducing the Intelligent Defense Platform

The detection gap in encrypted east-west traffic

Foundation models vs. analyst copilots: which does what

Vertical foundation models in production: what it actually takes

Slow credential abuse: detecting the attacks UEBA was built for but misses anyway