What Changed
A manual, time-intensive process for provisioning interview environments was replaced with a fully automated system that enables secure Azure sandbox creation and cleanup in a single click.
Where They Started
The company conducts technical interviews that require candidates to interact with real Azure resources such as storage accounts, databases, and app services. Previously, candidates were expected to provision these environments themselves, or administrators had to manually create and configure them through the Azure Portal.
What Was Breaking
The process was inefficient and inconsistent. Candidates often struggled to provision resources correctly, while administrators faced a time-consuming workflow involving manual resource creation, RBAC configuration, and post-interview cleanup.
This not only slowed down the interview process but also introduced risks of misconfiguration and orphaned resources accumulating unnecessary cloud costs.
How The Zig Fixed It
The Zig designed and built a full-stack system that automates the entire lifecycle of interview sandbox environments.
Initially developed as a desktop application, the platform was re-architected into a cloud-native web application to eliminate distribution limitations and enable access from any browser.
The backend, built using Azure Functions, handles candidate management, resource provisioning via ARM templates, Microsoft Entra integration for user access, and automated cleanup workflows.
The system dynamically provisions isolated resource groups, assigns RBAC roles, and ensures that all resources are tracked and removed after use, including safe cleanup of guest users where applicable.
What It Unlocked
Interview setup time was reduced from a multi-step manual process to a single action, significantly improving efficiency for interviewers and consistency for candidates.
The system also eliminated the risk of orphaned resources and reduced cloud cost leakage, while enabling scalable, repeatable technical assessments.
Where the Investment Went
Investment was concentrated in backend architecture, Azure resource management logic, frontend development, and Microsoft Graph integration. Additional effort was required to handle edge cases in resource provisioning and user lifecycle management.
What This Taught Us
Infrastructure automation is not just about speed—it’s about consistency and control.
This project reinforced the importance of designing systems that manage the full lifecycle of resources, including provisioning, access control, and cleanup, to avoid hidden operational costs.