What is intent detection?

Intent detection is the AI skill of figuring out what a person is trying to accomplish when they speak or type. If a caller says "I'd like to come in sometime next week," the intent is book an appointment. If they say "what are your hours on Saturday?", the intent is get business hours. Detecting intents accurately is what makes an AI voice agent actually useful — instead of just a transcription service with a voice on top.

Intent detection is a specific piece of the broader natural language understanding (NLU) pipeline. It's the decision point where the agent chooses what kind of action to take next.

How intent detection used to work

Traditional intent detection was based on keyword matching or, later, fixed-vocabulary classifiers. A developer would:

• Define a set of possible intents ("book_appointment", "check_hours", "speak_to_human", etc.)

• Label example phrases for each intent

• Train a classifier on the labeled examples

• Route callers based on the top predicted intent

This worked okay for a narrow set of intents in a narrow domain, and broke down the moment callers phrased things in unexpected ways.

How intent detection works now

Modern AI voice agents use large language models for intent detection. Instead of a fixed classifier, the LLM sees the caller's actual words along with the agent's instructions and figures out the intent contextually. The result:

• Works on any phrasing, not just ones seen in training data

• Can detect multiple intents in the same utterance ("book an appointment for Tuesday, and can you also send me a quote?")

• Can detect ambiguous intents and ask clarifying questions

• Adapts to changes in your business instructions without retraining

Intent detection vs. entity extraction

Intent answers "what do they want to do?" Entity extraction answers "what are the details?"

Example: "Can I book an appointment with Dr. Smith next Tuesday at 3pm?"

• Intent: book_appointment

• Entities: provider = "Dr. Smith", day = "next Tuesday", time = "3pm"

The AI needs both to complete the task correctly.

Why intent detection matters for your business

Every inbound call has one or more goals the caller is trying to accomplish. How well the AI detects those goals is the top driver of two key metrics:

• Call containment — the percentage of calls the AI handles without escalating to a human

• Caller satisfaction — whether the caller got what they came for without frustration

Good intent detection is why an AI agent feels responsive instead of robotic, and why callers stay on the line instead of pressing 0 to reach a person.

Related concepts

• Natural language understanding (NLU) — the broader pipeline intent detection is part of

• Large language model (LLM) — what powers modern intent detection

• Call deflection — the business metric intent detection drives

• Call containment rate — another business metric tied to intent detection

See it in action

The Receptionist Agent from 365agents uses LLM-based intent detection tuned for each customer's business. You describe what the agent should handle; it figures out how callers will ask for those things — no intent labeling required.