cpu memory allocation

I'm looking for a solution that can listen to Prometheus Alertmanager and Nginx and scale up Docker containers based on CPU/memory/traffic. It would be a solution that could scale the number of Docker containers up/down based on CPU/memory usage. Potentially via alerts from Alertmanager. It would need to work within a local Docker environment. It would need to perform scaling up/down based on CPU/memory usage and potentially Nginx traffic loads. So for example, if CPU use is over 80% for 10 minutes, the solution would add an extra container to help with load balancing. It would also scale down if traffic is low for a period of time. This could be a Flask app or similar, but essentially it's middleware between these stats and alerts and Docker