Engineering notes, research, customer stories, and the occasional rant from the team building autonomous pentesting at Intrudify.
Annual pentests are giving way to continuous, agent-driven testing. We unpack the architectural shift, what it means for security teams, and where humans still hold the edge over autonomous systems.
Read moreEvery Intrudify finding ships with a plain-English explanation, a suggested patch, and a confidence score. Here's how the LLM pipeline that produces them works - and why we built it after our customers asked us to.
Read moreThe deadline has come and gone, and most teams are scrambling. We mapped every NIS 2 article to an Intrudify control, plus the three tests auditors keep asking us to run first.
Read moreContinuous AI scanning catches the regressions; an annual human-led engagement catches the architectural risks. Here's how to budget for the two without either feeling redundant.
Read moreAtlas Fintech ran a side-by-side trial: a four-week engagement from a Big Four firm against six weeks of Intrudify runs. The findings overlap was 91%. The cost difference was 8×.
Read moreA walkthrough of the seven stages every scan goes through - discovery, parameter classification, payload synthesis, validation, scoring, remediation drafting, report assembly. Plus the SLOs that keep it under 24 hours.
Read moreAuditors expect different evidence formats from each framework. We built a single artifact that ticks all three boxes - and renewed our own SOC 2 with it last quarter.
Read more