# Hosting Options Aiimi Workplace AI can be hosted in several ways depending on your preferences and IT set up. ## On-Premises {% hint style="info" %} This requires servers or virtual machines that can host the right Windows Server or Linux. {% endhint %} The platform can be hosted on-premises in your data centre. This is best if all agent servers, repositories and web servers also run on local servers. *** ## Cloud The platform can be hosted in the cloud using any provider that offers PaaS. This is best if all agent servers, repositories and web servers also run on cloud-based servers. We can accommodate all of the main cloud platforms, Including but not limited to, Amazon AWS, Microsoft Azure and Google Compute Platform. Using cloud platforms can simplify scaling, especially for the repository tier. This is great if you need more Elasticsearch nodes as the volume of data grows, or demand increases. *** ## Hybrid The best solution if you have information on the cloud and on-premises servers. Putting source, content, and enrichment agents on-premises can improve the discovery and enrichment time. It is helpful for large on-premise sources such as SharePoint or network file systems. Some have on-premises agents for certain source systems and cloud agents for their cloud sources. Running the repository, API and User Interface in the cloud, can make scaling easy. It can also make end-user access easy since users do not need to be on the corporate network/VPN. *This depends on the security needed for Workplace AI and how users access the platform.* #### A common hybrid configuration * The agents are run on-premises. * The repository and web components are run on the cloud.