US-Based servers for reduced latency in live chat latency sensitive areas.
Hey, we use the service to quickly decide on whether to divert to our language translation model. Calling our language model unnecessarily for natively supported languages adds a lot of latency and is inefficient, so we enjoy having the ability to see ahead of time whether we should invoke it or not. Would it be possible to request a basic setup of some US based servers to improve the latency for our chat-sensitive applications (i.e. reduce the latency before a live chat message is posted)? Right now it adds about 100-200ms ontop of the 400ms it takes on our side, so it's approaching that barrier of being noticeable for our downstream customers.
This year we have deployed 4 new locations, including US East and US West. Requests are automatically routed to the closest data centre. It significantly reduces API latency for the clients worldwide.
Please find more information here.