Scaling and Orchestrating “Good Bot” With Kubernetes
summary
A Kubernetes-based orchestrator deploys containerized human agents that run javascript-enabled browser workloads, file transfers and app interactions, with a lightweight image able to scale from single digits to thousands of bots on modest nodes. Behavior models fit real user datasets (statistical tests show close distributional matches) and a command-and-control API provides real-time scaling, targeting and timing control; the system runs in testbeds and proves practical for large-scale replay of realistic normal traffic.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:50Cybersecurity training platforms
- 6:51User traffic simulation needs
- 10:11Designing Kubernetes-based bot
- 16:25Modeling user behavior
- 20:46Testing and deployment
- 24:42Simulating complex scenarios
- 28:47User mobility in simulations
- 32:50Collaborative efforts and Q&A
related talks
27mFrom Chatbots to Agents: Orchestrating Agentic AI on Existing Enterprise Kubernetes Infrastructure.
35mTo Infinite Scale and Beyond: Operating Kubernetes Past the Steady State
37mDeploying applications to Kubernetes: never send a human to do a machine's job
39mBuilding Distributed Systems with Kubernetes
42mAutomating the Enterprise with Java-based Kubernetes Operators
01h 01mDeploying Kubernetes, a Container Cluster Manager
19mDisaggregated Intelligence: Scaling AI Agents with Dynamo and Kubernetes
37mRunning Kubernetes in a Manufacturing Line