Design pipeline using classification and embedding services

Imagine you have access to two external, closed-source ML APIs: a classifier and an embedder. The classifier accepts a batch of text documents and returns a categorical label for each document; the embedder accepts the same batch and returns a fixed-length vector (e.g., 768 dimensions) per document. You are tasked with building a backend system around these services. The system must let users upload files, where each file may contain one or more text documents. It must…