Providing both robust security for your data and high performance for fast results, a range of deployment options are available for the Alchemite™ software. Compatible with Amazon AWS, Microsoft Azure, and Google, and supporting cloud or on-site models, these enable you to get up-and-running fast while fitting with your organisation’s IT policies and infrastructure.
Hosting options
Shared hosting: this hosting option is included in the standard user licence fees. Users log in to a secure, shared cluster, hosted by Intellegens. This is the fastest and easiest option to get started – we create an account for each of your users and you can be up-and-running immediately.
Dedicated cloud hosting: Alchemite™ can be deployed on a private cluster (Amazon AWS, Microsoft Azure, and Google are all supported), managed by Intellegens or by your IT team in their own cloud environment. Additional fees apply.
On site deployment: Intellegens can provide implementation support to internal IT Teams. For guidance, the system can operate on a minimum single node system of 32 GB of RAM with 8 cores and 250Gb filesystem storage. A recommended system would be multiple node (minimum 3), a minimum of 16 GB of RAM, with 4 cores, 100GB file system storage per node, plus 500GB for the File System and 3x100GB for OpenSearch datastore.
Technology background
The Alchemite™ platform is a containerised system, installed within a Kubernetes cluster. Kubernetes is a container orchestration system for automating deployment, scaling, and management of containers. This approach allows for compute resources to be automatically scaled up and down as required. High computational resources are needed to train models and when a large number of models or users are to be served.
Amazon AWS, Microsoft Azure and Google provide hosted Kubernetes environments, meaning Alchemite™ can be easily deployed across any cloud provider. It can also be installed on any internal compute cluster that can support Kubernetes.
Requirements for end users
Alchemite™ Analytics is a web browser based tool and all expensive computation will be done on the cloud, so if you’re using Alchemite™ Analytics there’s no requirement on your device/laptop. In order to get the best experience we recommend keeping your browser version up to date.
For Alchemite™ Engine API users, some familiarity with programming is required. We provide a Python client to simplify the process of authenticating with and using the API. Using this client requires a Python installation of 3.8 or newer. Alternatively, full documentation of the API is available to allow the creation of a custom client, if you have specific needs that the Python client can’t fulfil.
