Question

A retail company wants to build an internal tool where employees can upload a photo of a damaged product and receive a written summary describing the type of damage and suggested next steps. The team is deciding what kind of model capability they need. Which type of model is best suited for this requirement?

Accepted Answer

A multimodal model that can accept image input and generate text output

Answer

A text-only large language model that processes typed descriptions of the damage

Answer

A predictive machine learning model trained to forecast future return volumes

Answer

An embedding model that converts the photo into a vector for similarity search