Hands-on AI agent security workshop business with self-provisioned attack lab environments

A training product where attendees attack a deliberately permissive agent (a chatbot wired to a Kubernetes cluster with tools) and then repeat the same attacks against progressively hardened configurations — exposed, infrastructure-guarded, and AI-guardrail-enabled — learning what each layer stops and what it doesn't. Participants get their own disposable cluster, terminal, chat interface and observability backend via a self-service provisioning page, and the whole setup is open-sourced as a repo.

online BOTH Product / E-commerce

From AI Engineer — Build a Platform and Watch It Burn — Michael Forrester, Accenture & Whitney Lee, Datadog at 00:12

Problem: Teams letting agents into their clusters have no tangible sense of 'excessive agency' risk — a friendly chat interface becomes a route to tools, files and cluster operations — and no safe way to rehearse attacks and countermeasures before an incident.

For: Platform engineering teams, enterprise engineers, architects and C-level security stakeholders at companies adopting AI agents.

Products from this video

Amazon Elastic Kubernetes Service (Amazon EKS) AWS Bedrock Backstage BurritoBot Datadog Falco Falco Talon Grafana Jaeger kagent KubeArmor Kyverno LLM Guard OpenTelemetry SPIFFE/SPIRE Tetragon Watch It Burn - Claim Your Cluster Zipkin

Examples

🔒 Full analysis locked

Unlock more videos and the full analysis

A credit unlocks one video's full analysis for good — the build steps, the tools and how each was used, the methods behind every use case. Pro opens the whole library instead, and raises how many videos you can analyse a day.

Unlock full analysis — free

Other takes on Digital products

All Digital products ideas →