Use self trained model in the workflow

Hi,

I have done a workflow which categories play store reviews in pre defined categories. However, it does not work accurate all the time. And I was hoping if I can train the model using the past examples and then use the self trained model to categorize the new reviews.
I have tried creating separate workflows, one to train the model by reading existing data from Google spreadsheet and then using the same model through a webhook. Also, i have tried training the model first and then using it for categorizing. But since it does not retain information, it goes through all the existing past data for every new review and it gets too expensive to run.

Please suggest a correct way to do the same.

Regards
Nextgen.AI.Consult

Instead of retraining the model every time, use an AI service (like a custom model on platforms such as Google AutoML, AWS SageMaker, or OpenAI’s fine-tuning) where you train the model once with your past reviews. Then, just send new reviews for prediction only, so it doesn’t need to process all old data again.

hope this helps, cheers