AI Subnet Assistant — Plain English

100design patterns

12categories

0msnetwork latency

100%private (client-side)

Hi! I'm your network planning assistant.

Describe what you need and I'll generate a complete subnet design with reasoning. I recognize 100 distinct patterns across cloud architectures, Kubernetes, VLSM, IPv6, security segmentation, and more.

Try something like:

"3-tier AWS app in /16 across 3 AZs"
"K8s pod and service CIDRs for 5000 pods"
"Allocate 10.0.0.0/16 to teams of 500, 250, 100, 30 hosts"
"Office network for 200 employees with VLANs"
"How many hosts in /22?"

How it works: A pattern-matching engine runs in your browser, recognizes intent + extracts parameters (CIDRs, host counts, cloud provider, AZ count), and generates designs using deterministic CIDR math. No LLM, no API calls, no data leaves your device. Designs are guaranteed mathematically correct.

What the AI assistant does

Describe your network in plain English ("a 3-tier app in AWS with two AZs, ~200 app servers, ~20 databases, a small public ALB tier") and the assistant produces a complete CIDR plan with reasoning. It picks the parent VPC CIDR, decides subnet sizes based on your host counts, places them across availability zones, and explains its choices.

Under the hood it uses a 100-pattern rule engine — not an LLM call to a remote service — so your network description never leaves your browser. The patterns cover common architectures: web/app/data tiers, hub-and-spoke topologies, Kubernetes pod CIDRs, BGP peerings, and more. For designs the patterns don't cover, the assistant falls back to a reasonable default and explains what it inferred.

From description to deployable

The assistant's output is a structured plan. From there:

Refine the subnet sizes manually with the VLSM planner if you want different proportions.

Validate the cloud-provider reserved-IP math with the cloud-aware calculator.

Generate Terraform / CloudFormation / Pulumi via the IaC export tool.

Save the design to a workspace project for collaboration.

Background reading

For the architectural patterns the assistant recognizes, see multi-cloud CIDR planning and Kubernetes pod CIDR sizing. The private vs public subnets article explains the tier convention the assistant assumes.

Frequently asked questions

How does the AI assistant generate subnet plans?

The assistant uses a 100-pattern rule engine that matches your description against common network architectures — 3-tier web apps, hub-and-spoke topologies, Kubernetes pod CIDRs, BGP peerings, and others. It picks the parent VPC CIDR, sizes subnets based on your host counts, distributes them across availability zones, and explains its reasoning. The engine runs entirely in your browser.

Does the AI assistant send my network description to a remote server?

No. The assistant runs entirely client-side. Your network description, host counts, and the generated design never leave your browser. The 100-pattern rule engine is loaded as JavaScript and executes locally.

Can the AI handle non-standard architectures?

For architectures the patterns do not match exactly, the assistant falls back to a reasonable default and tells you what it inferred. For highly custom designs you may want to refine the result manually using the VLSM planner or calculator.

Plan networks in plain English