Google Cloud data migration platform

Scale without losing visibility

Educational institutions needed to move petabyte-scale datasets from on-premises environments into Google Cloud Storage. Workers handled the heavy processing, but the operation needed a secure way to configure, start, and monitor every migration.

The product could not treat transfer as a black box. Permissions, status, failures, duration, and cost had to remain understandable before and during execution.

The layer between people and workers

I was responsible for the front end and the proxy layer used to consume the internal API. This boundary organized authentication, response contracts, and communication with workers without exposing infrastructure details to the interface.

On the front end, React Query coordinated asynchronous state, status updates, and failure handling. Role-based access limited available actions according to each user’s responsibilities.

Planning before processing

One of the most valuable deliveries was the migration calculator. Based on volume and processing characteristics, the interface presented estimates for:

• duration in hours and minutes;
• required processing capacity;
• approximate cost in US dollars.

The estimate did not replace the actual outcome, but it turned an abstract infrastructure decision into a scenario that could be reviewed before execution.

Outcome

The platform brought configuration, permissions, tracking, and cost planning into a single workflow. Teams gained predictability for running extensive migrations and a clear view of what the workers were processing.

The case reinforces an idea that appears throughout my work: cloud interfaces are better when they translate infrastructure without hiding its behavior.