Quick question about the Quality Management API behavior during high-volume evaluation imports for our multi-org integration.
We are currently processing evaluation data via the POST /api/v2/quality/evaluations endpoint. While single-threaded requests succeed, our automated bulk import process triggers a 429 Too Many Requests error when the throughput exceeds approximately 15 requests per second per organization.
The documentation references rate limits, but it is unclear if these limits are aggregated across all OAuth tokens for a single AppFoundry application or if each client ID receives a distinct quota. We are operating under the assumption that rate limits are scoped to the application level, which creates a bottleneck when scaling to 50+ concurrent organizations.
{
"code": "tooManyRequests",
"message": "Rate limit exceeded for api:quality:write"
}
Has anyone successfully implemented a dynamic backoff strategy that respects these specific thresholds without hardcoding limits? We want to ensure our integration remains compliant and efficient across the entire partner ecosystem.