Configure The Development Environment

Ensure that you configured the host system according to Configure The Production Environment.

The Savant framework requires a complex environment, including GStreamer, Nvidia DeepStream, OpenCV CUDA, and other dependencies. Installing and configuring all the required packages is a daunting task. Therefore, we prepared ready-to-use docker images both for development and production usage. The image provides you with a fully configured environment in minutes.

Development Environment Problem

When we develop the code, we expect IDE’s assistance on available classes, their methods, properties, etc. The Savant codebase is in Docker, which makes a problem for the IDE because it cannot collect the database for symbols.

Normally, we use VENV to install packages, and IDE knows how to gather symbols from VENV; however, the Savant environment is too complex to install in VENV.

Fortunately, popular Python IDEs like PyCharm Professional and VSCode support setting up a development environment in a docker container: it gives the IDE information from where to collect available packages and their symbols.

Configuration Manuals

We will guide how to configure PyCharm Professional and VSCode to use runtime environment in Docker. The GitHub repository contains a module template; we will use it to demonstrate configuring the development environment in PyCharm Professional and VSCode.


PyCharm Community Edition doesn’t support development in the Docker container environment, so it cannot be used easily. If you consider only free-of-charge IDEs, consider VSCode.


Recommended spare disk space is 20 GB.