The adoption of cloud computing is continuously increasing due to the attractiveness of low costs of infrastructure acquisition and maintenance, as well as having virtually infinite resources available for scaling applications based on demand. Due to the increasing interest in this topic, there is a continuos search for better, more cost-effective ways to manage such infrastructures. One of the most recent steps was taken by the definition and development of Serverless computing, aka Function-as-a-Service (FaaS). FaaS is a cloud computing service model where developers can deploy functions to a cloud platform and have them executed based either on the triggering of events by other services, or by making requests directly to an HTTP (S) gateway, without having to worry about setting up the underlying infrastructure. In this paper, we propose an architecture for deploying FaaS platforms in hybrid clouds that can be composed by multiple cloud providers. This architecture aims at enabling privately deployed FaaS platforms to perform auto-scaling of resources (virtual machines) in a distributed infrastructure, while considering the scenario where the users of such platform are scattered around the globe. This allows the execution of requests in servers geographically located as close as possible from the client, with benefits to both the clients and the service providers.