SA Studio is the graphical user interface for SA Engine. With SA Studio you can develop and deploy models, and manage edge devices. It is an online application that you access through a regular browser on your workstation. From SA Studio you can access and control any device that runs SA Engine.
There is a license-free SA Studio Community Edition that is free to use for non-commercial applications. It comes with some restrictions but provides a convenient way of trying out analytics on edge devices.
SA Engine is a streaming analytics engine which operates on data streams and facilitate transformation of streams using computational and AI models (SA Studio also runs an instance of SA Engine). SA Engine can be installed on many different devices, such as microcontrollers, Raspberry Pis and PCs.
SA Studio has a versatile IDE in which the user can develop computational models, read documentation, connect edges to servers, etc.
From SA Studio the user can also:
- Visualize data streams from edge devices.
- Deploy models to edge devices.
- Interactively update models on edge devices.
And all this while the edge devices are running. There is no need to turn off or restart an edge device between deployments or model updates.
For model development, deployment and orchestration SA Engine uses a powerful query language called OSQL (Object Stream Query Language). It provides an experience that best can be described as a combination of SQL and Matlab. Once you get familiar with the syntax and capabilities of the language, developing and deploying computational models is easy and low effort compared to more traditional languages such as Python and C. It is designed to be easy to use for data scientists, engineers and analysts, without any need for deep programming skills.
SA Studio contains powerful tools for real-time visualization. This means you can visualize data streams, sensor values, and streaming model results in real time.
SA Engine also supports a variety of third-party technologies such as MQTT and Kafka, for sending data as messages to external systems, and TensorFlow Lite for inference with artificial neural networks.
SA Studio is a streaming analytics platform GUI that enable processes to work with data and develop models based on data and data streams. No user data or user created data is saved and stored unless a user explicitly uses the tools in the platform to do so. No data is stored in any external storage unless a user explicitly send the data there.
When using an edge device, no processing is done with the device sensors unless a user explicitly run queries requesting data streams. The data never leaves the edge device and is not stored, unless a user explicitly run a query sending the data to the platform for usage or storage.
The server of the platform will periodically go into sleep mode whereby data will not be stored unless a user chose to do so. An exception is all files stored under the menu "your models" that will be kept so a user can resume work.
Accessing SA Studio Community Edition
You can access SA Studio Community Edition by going to https://studio.streamanalyze.com. You need to register a user account by providing your email address. Once you have signed up you can start SA Studio from the same page.