Getting started with Snitch Hybrid or On-Premise
The Hybrid Mode and the On-Premise Mode (installed on organizations' or users' computers) offerings are available to users who wish to keep their datasets and models isolated within their own environments. With the Hybrid mode, only the analysis results are sent to Snitch to be centralized and displayed in our online dashboard. With the On-Premise mode, nothing is sent to our servers at all, leaving the user to manage the analysis results as they see fit. With both Hybrid and On-Premise modes, datasets and model files are never sent to Snitch's servers.
Which environment is right for you?
| Analysis results available in PDF and JSON formats
| Access to Python Library
| Online dashboard to view and compare analysis results
| Can launch analyses via the dashboard
| Local installation required (container image)
|Your datasets and models never leave your infrastructure||❌||✅||✅|
Installing the Snitch agent
To use Snitch Hybrid or On-Prem, you must use a container orchestrator (such as Docker Desktop) to host our container image. Once the image has been pulled from our repository, it must be configured and can then serve API requests.
We give you step-by-step instructions on how to install and use Snitch with Docker Desktop. However, you can choose your own container orchestrator if you so desire. If you do, the commands may vary to achieve the same result.
Step 1: Install Docker Desktop
- Download Docker Desktop from this website: https://www.docker.com/products/docker-desktop
- Install Docker Desktop on your machine. If you are installing on Windows, prefer WSL2 integration for better and more reliable performance.
Step 2: Pull the image from our repository
- Login to our repository using your credentials.
docker login -u <repository username> -p <repository password> snitchai.azurecr.io
To view the credentials to the Docker registry, click on "Show Docker credentials" in the user menu on the top-right part of the screen:
- Pull the latest image from the repository.
docker pull snitchai.azurecr.io/snitch-agent:latest
Step 3: Start the agent
- Create a Docker volume
docker volume create snitch-storage
- From the Snitch app at https://app.snit.ch, click on "Manage access tokens" in the user menu on the top-right part of the screen. Generate an access token and make sure to note the value. You will no longer be able to view the token once the window is closed.
- Run the Snitch Agent container
docker run -dt --name my-snitch-agent --mount source=snitch-storage,target=/etc/snitch -p 5443:443 -e SNITCH_ACCESS_TOKEN=<your_access_token> snitchai.azurecr.io/snitch-agent:latest
- [On-Premise mode only] Add the following to the command-line above:
Step 4: Use the Python API to use the Snitch agent
- Install the Snitch AI Python package
pip install snitch_ai
- Usage instructions and examples are available on our PyPi page: https://pypi.org/project/snitch-ai/
- If you followed our instructions exactly and the agent is installed on your workstation, the endpoint_address value will be https://localhost:5443/
You're now good to go! If you run into any issues or require assistance, please feel free to contact our support team who will be more than happy to assist you.