What is the standard approach to handle Conversation API rate limits during high-volume bot handoffs? Our AppFoundry integration manages a significant volume of digital interactions, and we are encountering 429 Too Many Requests errors on the /api/v2/conversations/webchat endpoint when scaling up concurrent sessions. The issue manifests specifically when the AI bot triggers a complex handoff logic that requires multiple API calls to update participant roles and transfer ownership within a sub-second window. We have implemented standard exponential backoff, but the latency introduced causes session timeouts on the client side, resulting in dropped conversations.
The environment is using the latest Genesys Cloud SDK for Node.js (v6.14.2) and operates across multiple organizations using OAuth 2.0. We suspect the rate limiting is hitting the per-org threshold rather than the global limit, but the documentation is unclear on how to distinguish between the two in error headers. Is there a recommended pattern for batching these updates or a specific header to request higher throughput for partner applications? We need a reliable method to ensure data consistency without disrupting the user experience during peak load periods.