The complete learning path

Orientation, and your first lesson: what network automation actually is.

The core concepts behind how network automation actually works: inventory, state, idempotency, and safe failure handling.

Just enough Python, taught with networking examples throughout — variables, data structures, loops, functions, and exception handling, ending in a real automation script.

Reading and writing JSON, YAML, and CSV — the formats that carry almost all real network automation data.

Build a complete network health check end to end — inventory, a simulated connection layer, validation, and a real report.

Connect to real devices over SSH with Netmiko — a real, read-only health check, built on the project from Your First Network Automation.

Retry policies, pre/post-check validation, dry-run mode, and structured results — making the Netmiko Fundamentals project genuinely reliable.

Structured data, rendered through Jinja2 templates, into reviewed and validated configuration — extending Reliable Automation's safe-deployment pipeline.

Declarative, agentless automation — the same VLAN deployment problem from Configuration & Template Automation, now orchestrated with real Ansible inventory, playbooks, and check mode.

The HTTP request/response model, the requests library, JSON responses, and safe read-before-write API automation — the same VLAN health-check lineage since Phase 2.6, now over HTTP instead of SSH.

Controller-centric automation — centralized inventory, asynchronous tasks, and observed-state validation, illustrated through a vendor-neutral mock controller and a real Cisco Catalyst Center example.

From one-off script to reusable tool — dataclasses, custom exceptions, classes, logging, unit testing, bounded concurrency, and a real CLI, built around the automation capabilities already learned.

Python-native, inventory-driven automation — the framework version of the inventory, task execution, and concurrency built by hand in Advanced Python Automation, plus filtering and a nornir_netmiko integration.

Vendor-neutral network state -- normalized getters (get_facts, get_interfaces, get_interfaces_ip), a real candidate-configuration workflow, and NAPALM's own built-in mock driver for testing without real devices.

Nornir orchestration + NAPALM normalized state, combined through nornir_napalm -- one workflow running unmodified across a mixed-vendor fleet, with driver selection resolved automatically from inventory.

Inventory as external, structured data -- pynetbox as a real REST client, safe reads and writes, and a real Nornir inventory plugin that builds your fleet dynamically from NetBox instead of a static YAML file.

Version control for network configuration and automation artifacts -- real Git and GitPython, a branch-diff-merge review workflow, and versioning rendered configs and source-of-truth snapshots over time.

The bridge from an already-reviewed change to a safe production rollout -- staging blast radius with nr.filter(), a real threshold gate that decides expand or abort, and automatic rollback when a stage fails.

Three end-to-end projects combining previously-separate modules -- config backup with drift history, a multi-vendor compliance report sourced from NetBox, and a Git-reviewed change driving a staged, gated rollout -- plus guidance for scoping and presenting a project of your own.

One cumulative system, not a new tool: an NTP server migration, from a NetBox-sourced fleet through detection, generated remediation, Git evidence, and a staged, gated rollout with rollback -- closing the original curriculum roadmap.