Need some troubleshooting help with a weird integration issue between our WFM schedule publishing workflow and the new AI Virtual Agent capacity planning logic. We are running the latest WFM API version (v1) in the Chicago cluster, and everything looks fine on the surface, but the bot utilization metrics are completely out of sync with the actual schedule adherence data. When I trigger the weekly schedule publish for our Tier 2 support team, the WFM system correctly marks agents as ‘Available’ for digital channels, but the AI Orchestrator continues to route inbound intents to the human queue instead of the bot, even though the bot’s confidence threshold is set to 0.85. I dug into the Architect flow logs and noticed a specific error code 429 Too Many Requests appearing in the WFM status webhook endpoint right after the schedule goes live. This seems to cause a delay in the status propagation to the routing engine. The agents are showing as available in the WFM dashboard, but the real-time capacity view in the Admin console still shows zero digital availability for the first 15 minutes after publish. Is there a known latency issue with the WFM-to-AI capacity sync in the US East region? I have checked the API rate limits and we are well within the 50 requests per second limit for the WFM endpoints. The bot flow itself is working fine in isolation, so this definitely seems like a data sync problem rather than a flow configuration error. I am trying to avoid having to manually refresh the routing profiles every Monday morning when the new schedules drop. Any insights on how to force a faster sync or if there is a specific header I need to include in the WFM publish call to trigger an immediate update to the AI capacity engine? This is causing a significant bottleneck in our shift swap approvals because agents think they are covered by the bot when they are not, leading to missed SLAs during the peak morning rush.