AI / Voice
NLU
Natural Language Understanding (NLU) is the capability that extracts meaning, intent, entities, and context from a person's spoken or written language.
Natural Language Understanding (NLU) is the capability that extracts meaning, intent, entities, and context from a person's spoken or written language.
Natural Language Understanding, or NLU, is the capability that turns language into usable meaning. It identifies what a person wants, the important details in the request, and how the current message relates to earlier turns. In a phone conversation, NLU works on text produced by speech recognitionSpeech-to-Text (STT)Speech-to-Text (STT) converts spoken audio into text that a voice agent can interpret, store as a transcript, and use in downstream actions. rather than on the raw audio itself.
What NLU identifies
Common outputs include intent, entities, sentiment signals, and structured values. A caller might say, "I need something next Tuesday after lunch." NLU can identify scheduling intent, the requested day, and a time preference even though the person did not use a fixed command. Modern systems may perform this work with dedicated classifiers, an LLMLarge Language Model (LLM)A large language model (LLM) is a machine-learning model trained on large text collections to interpret and generate language from prompts and context., deterministic rules, or a combination.
Practical use cases
- Distinguish a new inquiry from a rescheduling or cancellation request.
- Capture location, timeline, product interest, or eligibility details.
- Route a caller based on the stated need and collected context.
- Recognize when an answer is unclear and ask a focused follow-up question.
NLU should not silently turn uncertain language into a high-stakes decision. Thoughtly's outcomes support prompt-based interpretation for open conversation and rule-based checks for structured validation or critical logic. That division is useful: flexibility for language, explicit control for consequences.
Frequently asked questions
Is NLU the same as speech-to-text?
No. Speech-to-text records what was said. NLU interprets what that text means in the current conversation.
Is NLU the same as an LLM?
No. An LLM can provide NLU capabilities, but NLU describes the function. It may also be implemented with classifiers, rules, or other models.