ai-integration-examples | integrating external AI services into Sisense

Developing new language model integrations with Sisense can be done in one of the two following ways:. The default behavior is Remote execution. To run it locally, the code changes listed below are required.
The "Remote" option - Via the Jupyter server in your Sisense instance.
The "Local" option – Execution via local IDE.
This type of installation depends on your deployment type:.
On-prem customers should refer here for a terminal based installation
Cloud managed customers and on-prem customers who prefer a UI based installation should refer here (Sisense version 23.2 and above)
Log into Sisense and enable Custom-Code, Custom-Code Transformations and Blox as shown here
Clone this repository.
Edit the configuration file.
Open your terminal and cd into the repository's directory.
Run the following command:
Clone this repository.
Install the required libraries using the following command:
Extract authentication token from Sisense REST API explained here
Uncomment the code and Insert OpenAI API token here
Change log file destination path to local repo uncomment here, comment here
Set the following environment variables: API_GATEWAY_EXTERNAL_SERVICE_HOST={Sisense instance IP} API_GATEWAY_EXTERNAL_SERVICE_PORT={Sisense instance port} Cookie={Sisense authentication token} (can be done within the notebooks)
When executing a notebook, pay attention to the required changes listed in the section How to run the notebook
You can now run the notebooks in the repo or create a new one
Creating a new notebook requires additional steps in order for it to be recognised by Sisense, please refer to Adding a New Notebook
You should also upload the utils files to the server at http://{Sisense IP}:{Sisense Port}/app/diag/tree/work/storage_notebooks/custom_code_notebooks/notebooks