01Overview
Hands-on program for detection engineers who want to integrate AI primitives into pipelines without losing analytical discipline.
What you will do
- 01Apply embeddings to telemetry triage at scale
- 02Build LLM-assisted alert summarization that survives audit
- 03Recognize when not to use AI
- 04Measure detection quality under AI augmentation
- 05Defend your design under operational scrutiny
The scenario
We work from real telemetry datasets and walk through the failure modes of naive AI integration.
Why this exists
AI in detection is full of dangerous shortcuts. We make the trade-offs explicit.
- Jul 1Day one opens141 days ago
- Jul 3Last day139 days ago
- Nov 19Bootcamp publishedtoday
- Level
- Intermediate
- Labs
- One sandbox, one per lab, started from the page in a browser terminal on an isolated network, 15 to 240 minutes with up to 60 more
02Who it is for
Detection engineers and threat hunters with pipeline experience.
Outcomes
- 01Use embeddings to compress noisy telemetry.
- 02Design LLM-assisted analyst workflows that don't degrade rigor.
- 03Articulate when AI hurts more than it helps.
- 04Measure detection quality changes under AI augmentation.
- 05Build operational guardrails.
03Syllabus, day by day
The detail of this week is published before the cohort opens.
The detail of this week is published before the cohort opens.
The detail of this week is published before the cohort opens.
The detail of this week is published before the cohort opens.
Capstone
Ship a detection augmentation pipeline that survives a hostile review.
04Instructor
Guest Engineer
Detection engineering lead, applied ML practitioner.
Talks on responsible AI in detection.
Years of detection engineering at scale.
05Schedule and cohort
- Day one
- Jul 1
- Last day
- Jul 3
- Enrolment
- Open
06Live sessions
Nothing on the calendar yet.
07Tuition and questions
Questions
Do I need ML background?
Helpful but not required.