A global e-commerce company hosts its application on EC2 behind an Application Load Balancer in us-east-1, with CloudFront in front. Users in Southeast Asia have started reporting slow page loads and intermittent errors over the past two days, but internal synthetic canaries running from AWS regions show healthy latency. The network operations team needs to determine whether the degradation is caused by internet path issues (such as an ISP or transit provider problem) affecting specific city/ASN combinations, and quantify the impact on real users without deploying agents. Which AWS capability should the team use?