We operate our entire AWS infrastructure in us-west-2 — S3 buckets, EKS clusters, data warehouse, and all downstream consumers. Our Fivetran Delta Lake destination is also configured for us-west-2.

However, since Polaris workers only run in us-east-1, every byte Fivetran writes to our S3 bucket crosses regions. This creates two unnecessary cost centers on our AWS bill:

1. Cross-region data transfer: Significant monthly USW2-USE1-AWS-Out-Bytes charges, purely from Fivetran's Polaris workers in us-east-1 accessing our us-west-2 bucket.

2. Cross-region API request overhead: We traced via CloudTrail that our Fivetran role generates tens of thousands of AssumeRole calls per day from us-east-1, all with the hll/cross-region user-agent flag — confirming the workers are operating cross-region.

These costs are entirely avoidable if Polaris workers were co-located with the destination. We can't move our destination to us-east-1 because the rest of our stack is in us-west-2 — moving the bucket would just shift the cross-region costs to our own read path instead.

Request: Deploy Polaris workers in us-west-2 (and ideally all regions where S3 destinations are supported) so that data processing happens in the same region as the destination bucket.

Business impact:

• Would immediately eliminate hundreds of dollars per month in AWS cross-region transfer fees for us
• Would reduce S3 API latency for Fivetran syncs (same-region calls are faster)
• Likely affects all customers whose Delta Lake destinations are outside us-east-1
• Aligns with the destination's configured "Cloud region" setting in the Fivetran UI, which currently says us-west-2 but doesn't actually control where Polaris workers execute

Evidence provided in support ticket: CloudTrail logs showing sts:AssumeRole from us-east-1 IPs, hll/cross-region user-agent flags, and S3 data transfer billing breakdowns.