AS Alan Scholl
CML Lab · ON-DEMAND Kept stopped to keep cloud cost near zero — available for walkthrough by request. Last updated 2026.

Home Lab: AI-Driven Cisco Modeling Labs on GCP

A nested-virtualization lab factory for CCNA practice, AI automation, and Infrastructure as Code — planned with ChatGPT, built with Claude Desktop, and deployed on Google Cloud.

An on-demand CCNA lab factory in the cloud

Instead of buying rack-mount lab gear, I run enterprise-grade network simulations on Google Cloud using hardware-assisted nested virtualization — spinning up multi-vendor topologies in minutes and tearing them down to near-zero cost when idle. It demonstrates practical skills in Infrastructure as Code, nested virtualization, cloud cost engineering, and AI-assisted planning and deployment.

Planned with

ChatGPT

Solution design and lab-scenario planning.

Built with

Claude Desktop

Coding, orchestration, and Infrastructure as Code.

Deployed with

Google Antigravity

Google's agentic, AI-assisted deployment tool for Google Cloud — used here for deployment and operations acceleration.

View the Presentation PDF Read the Full Project Write-Up Request a live demo

System architecture

Five layers, from AI planning down to live virtual routers — every layer defined as code.

stack> User/AI → Automation (Terraform·YAML·Python) → Google Cloud (VPC·Firewall·VM) → KVM/QEMU (nested) → Lab (IOSv routers, VLANs, STP)
# hardware-assisted nested virtualization runs a hypervisor inside a cloud VM
  • User / AI layer

    AI planning plus an engineer review loop drives requirements and iterations.

  • Automation layer

    Terraform (IaC), YAML configuration, and Python REST API orchestration via virl2_client.

  • Google Cloud layer

    Dedicated VPC, least-privilege firewall rules, and a Compute Engine controller VM.

  • VM / hypervisor layer

    n2-standard-4 (4 vCPU / 16 GB) on Intel Cascade Lake with nested virtualization enabled; KVM/QEMU runs inside the cloud VM.

  • Lab factory layer

    On-demand IOSv routers/switches, VLANs, STP, and injected hidden break/fix faults for troubleshooting practice.

Core stack

  • Google Cloud Platform
  • Cisco Modeling Labs
  • Terraform
  • YAML
  • Python REST API (virl2_client)
  • Nested virtualization (KVM/QEMU)
  • Intel Cascade Lake
  • ChatGPT
  • Claude Desktop
  • Google Antigravity

Cloud cost engineering

An ephemeral, on-demand lifecycle keeps the lab cheap: the controller is started for a session and stopped when idle, cutting monthly cost dramatically while every lab design stays saved on disk.

~$4.81Per day while the lab is running
~$0.24Per day when stopped (disk kept)
~61%Lower monthly cost by stopping when idle

Mapped to CCNA objectives

This lab is the hands-on engine behind my active CCNA sprint — each exam domain has a place to practice.

Network Fundamentals

Topology planning, addressing, virtualization context.

Network Access

VLANs, trunks, STP, switch behavior.

IP Connectivity

Router-on-a-stick, reachability, routing validation.

IP Services

Management access, automation APIs, service readiness.

Security Fundamentals

Firewall rules, controlled access, hardening considerations.

Automation & Programmability

Terraform, Python APIs, YAML, AI-assisted workflows.

See the 90-day cert sprint

Request a Lab Walkthrough

The lab is on-demand to keep cloud costs near zero. Request a walkthrough and I'll bring the environment online and walk you through it live. Here's what a reviewer can see:

  • CML network topology and injected break/fix troubleshooting scenarios
  • Terraform + Python (virl2_client) automation workflow
  • On-demand cost-control lifecycle (start/stop, disk-only idle state)
  • Related: the AI-driven Proxmox automation capstone (separate project)
Request a live demo View the Presentation PDF