Infrastructure as Code: Theory and Concepts

The talk makes the case for managing infrastructure through code instead of manual configuration. The presenter lists the practical gains: faster and repeatable deployments, a version-controlled record of changes that doubles as documentation, and simpler rollback. Infrastructure is defined broadly, covering network devices, servers, load balancers and other connected systems that can be managed programmatically.

The middle section connects software design to infrastructure work. Object-oriented ideas such as code reuse, loose coupling and single responsibility are applied to deployment scripts, with a class-based VPC example compared against a procedural approach. Factory and builder patterns are presented as ways to move from large monolithic templates toward modular environments. Operations are mapped to create, read, update, delete and upsert, which then become API calls. Postman and vendor SDKs are shown as ways to generate usable code, and configuration management and cloud-native template tools are surveyed.

The closing part connects these practices to CI/CD and DevOps. Pipelines add build, test and gated delivery stages, and version history gives auditors a record of changes, including when and where they were applied. The presenter argues that DevOps is mainly about organizational process and collaboration rather than tools, and recommends starting with one small target, such as a single firewall, before expanding. Audience questions covered winning over reluctant teams and applying these ideas where duties are segregated.