Customer support · Production
Large Language Models and Where to Use Them: Part 2 — Search, Cluster, and Classify (Cohere Blog)
The problem
The volume of unstructured text data being generated is too large for humans to process without NLP-powered automation.
Workflow diagram · grounded in source
1
Text embedding generation
ai_action
“the model generates a set of numbers that represent the meaning or context of the input text. These numbers are called "text embeddings"”
2
Semantic similarity search
ai_action
“given a question, the search engine would return other frequently asked questions (FAQ) whose text embeddings are the most similar to the question”
3
Document clustering
ai_action
“clustering them into smaller groups by the theme or topic of the posts, supplemented by the keywords that represent the topic of each group”
4
Text classification
ai_action
“using the Classify endpoint to perform sentiment analysis, which is a classification task that classifies a piece of text into one of the Positive, Neutral, and Negative classes”
Reported outcome
(not stated)
Reported stack
EmbedClassify
Frequently asked questions
What did this team achieve with this AI workflow?
(not stated)
What tools did this team use?
Embed, Classify.
How is this customer support AI workflow structured?
Text embedding generation → Semantic similarity search → Document clustering → Text classification.