Your company's product team has introduced a new requirement, prompted by customer demand, to implement autoscaling for your stateless and distributed service operating within a Google Kubernetes Engine (GKE) cluster. Given the tight timeline of two weeks before this feature needs to be live, what would be the most suitable approach?