Once you have created an AI experiment and run the model, as described in Working with AI experiments, if you are satisfied with its results, you can publish it so it can be used in a Production environment.
Before you continue
Publishing an experiment means that you are completely sure that the accuracy or standard error of the results produced by that experiment are satisfactory.
Before publishing, consider:
1. Publish experiments whose models have been generated against production data (targeting a production-environment Dataset).
2. Publishing an experiment makes it available for use in a Production environment, so you will not be able to modify or delete the experiment after you publish it.
3. Maintain a published experiment by periodically training/generating the model.
As is common in machine learning capabilities, you need to manage the learning curve of your model to make sure it reaches optimal behavior.
Note that even though you may have your Bizagi processes constantly feed the source Dataset, the AI experiment will not change or act upon new information unless you explicitly use its Run model option.
4. You should make sure that the data is final and doesn't need further preparation once it resides in the Dataset.
Using Run model frequently will help you train the model so that it is more reliable; however this does not guarantees that your data is exempt from having high variance or bias.
It is recommended to have large amount of information in the Dataset to produce reliable experiments.
To publish an experiment, follow these steps:
1. Click the Publish button.
Do this where you see the experiment's results:
Bizagi Artificial Intelligence shows a confirmation message to make sure you want to publish this prediction. You can only publish one experiment per model. Publishing a new experiment based on the same model results in un-publishing the previous one and projects will no longer be able to use it.
Click the Publish button.
For a published experiment, you can configure Rules settings or look up the connectivity information by clicking the View connectivity info button.
2. Consider Testing Data if necessary.
At this point, your AI experiment has been published to both a testing and a Production environment (each with a different URL).
As a last, optional step, you may choose to include rules to apply when using the Testing service.
To do this, click Add rule button at the end of the page.
In these settings, you can:
•Add any number of rules, which can use any number of conditions.
The conditions (e.g, the number of times pregnant is greater than or equal to 12) can directly influence the prediction.
•Choose a default value for all predictions.
Details about the HTTP request information, URL, access keys, and the connector are not relevant at this point to complete the publishing procedure.
Now that your experiment has been published, your Bizagi processes can rely on its service.
To learn how to do this, refer to Connecting Bizagi processes to use AI.