How would you design an API gateway for Select Minds to handle high-traffic spikes while ensuring low latency and high availability?
