Native, High-Throughput Integration of Concert Workflows as an Automation Target in CP4AIOPS

Our organization leverages IBM Cloud Pak for AIOps (CP4AIOPS) and needs to route deduplicated Alerts and Incidents to ServiceNow in TMF format via GCP PubSub. To optimize performance and reduce complexity, our ideal architectural pipeline aims to minimize network hops while leveraging modern low-code platforms: 
CP4AIOPS >> Concert Workflows API >> GCP PubSub >> ServiceNow
Current Limitation & Bottleneck 
Currently, invoking the Concert Workflows API directly from CP4AIOPS is heavily bottlenecked. Our environment is dimensioned for 650 raw events per second and in the worst case, we expect a similar throughput to egress out of CP4AIOPS. 

To process this volume today, we are forced to use IBM Impact as an intermediary routing engine. Impact is capable of handling ~450 calls per second because it can be clustered and scaled horizontally. This forces us into a suboptimal architecture (CP4AIOPS >> Impact >> Concert Workflows >> GCP PubSub >> ServiceNow). While Impact provides the necessary throughput, it introduces an unnecessary architectural hop, increases processing overhead, and delays our transition to a streamlined, workflow-centric integration model.

Proposed Enhancement 
We request native, high-capacity support within CP4AIOPS (Concert Operate) to treat Concert Workflows directly as an automation target. The feature should include:

Scalable Throughput: The direct CP4AIOPS-to-Concert Workflows integration must match or exceed the performance capabilities of clustered Impact servers, comfortably handling 650+ Invocations /sec without queuing or degradation.

Native Configuration: Provide out-of-the-box UI configuration to select Concert Workflows as a primary target, removing the need for custom development cycles or intermediary routing policies to manage payload bottlenecks.

Horizontal Scalability: Ensure the API invocation mechanism between AIOps and Workflows scales dynamically in an OpenShift (OCP) environment to prevent latency.