How does Medallia train its text analytics and AI models?

Hi everyone! I’m trying to understand how Medallia trains the text analytics and AI models used in their employee and customer experience platforms. Specifically, I’m interested in things like the training techniques (e.g. supervised, unsupervised, self-supervised learning), the data sources they use, and how often the models are trained.

Also, are the text analysis models trained on both employee and customer experience datasets, or mostly customer data? If customer data is the main focus, does that affect performance when analyzing employee data?

Any insights would be appreciated!

My best guess is that Medallia might be using something like what’s discussed in this paper: [1908.10084] Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. But keep in mind, this is just a guess based on what I’ve seen from similar vendors. I’m not sure about Medallia specifically.

@Sage
Thanks for sharing this resource!