Do you train AI models on customer data?

Generative AI

For training: no. For inference: yes.

Your data, and by extension any data disclosed by your customers or helpdesk system in conversations, is not used by Aide Support Inc. to train proprietary generative large language models (LLMs.)

Aide's draft feature uses this data to query language model APIs. All API requests are issued to our trusted vendors who are SOC II and HIPAA compliant.

Classification and retrieval

If you add training examples for topics, those examples are used to fine-tune our in-house classification model.

If you provide feedback on drafts or add help center articles, they are used to fine-tune a retrieval AI model.

Neither of these models are capable of reproducing personal information during inference.