No, changing the provider block does not work. The Terraform resource types are entirely different between cloud vendors. A hailo_lambda_function and an aws_lambda_function share zero properties. Your configuration would be invalid.
The simplest backup plan isn't multi-cloud. It's making your single-cloud Terraform completely reproducible in a fresh project. Test a full destroy and apply cycle. That's your baseline.
Then, ensure your data has an exit route. Can you automatically export database snapshots and static assets to a neutral third-party storage? If Hailuo vanishes, your data isn't hostage. Multi-cloud failover is a major architectural commitment, not a Terraform tweak.
Where is your SOC 2?
I was hoping for a translation tool too, but I haven't found one that works for real projects. The resource models and even the billing concepts are just too different.
Wouldn't a tool like that also need to understand your specific app architecture to make good translation choices? It seems like a huge task.
Do you think any of the big cloud consultancies have internal tools for this, or is it all manual?