Deskhero's AI capability 'suggested replies' works by analyzing previous tickets, knowledge base articles, and other relevant data related to the manufacturing of insulation materials. It generates embeddings vectors for each ticket, article, or scraped website, which are then stored in a vector database. When a new ticket is created or a search query is entered, the AI runs the query against the embeddings API and searches for similar relevant data. The relevant data is then passed to ChatGPT, where it generates a response based on the attached data. This response is presented as a suggested reply for support agents to use.
Use case 1
A support agent receives a ticket about a specific issue with mineral wool insulation. Deskhero's AI suggests a reply based on similar tickets in the past, providing a solution that has worked before.
Use case 2
A support agent searches for information about the fire resistance of insulation materials. Deskhero's AI analyzes the available data and suggests a reply that explains the fire resistance properties of mineral wool and other insulation materials.
Use case 3
A support agent receives a ticket asking for recommendations on the best insulation material for a specific application. Deskhero's AI suggests a reply based on previous tickets and knowledge base articles, providing a list of suitable insulation materials along with their benefits and drawbacks.
