1. Home
  2. Glossary
  3. Natural language processing (NLP)
AI glossary · Foundations

Natural language processing (NLP)

Natural language processing (NLP): Natural language processing (NLP) is the branch of AI focused on understanding and generating human language: reading, classifying, translating, summarizing, and answering questions in ordinary words. Large language models are its current dominant technology.

For decades, getting software to handle language meant hand-built rules and statistical tricks: spell checkers, spam filters, keyword search, basic sentiment scoring. Each task needed its own specialized system, and results were brittle whenever people phrased things in unexpected ways.

Large language models changed that by handling most language tasks with one general model. Classification, extraction, translation, summarization, and drafting are now things you ask for in plain English rather than build. Many older NLP techniques still run in the background of products you use, such as search and voice assistants, because they are cheap and fast.

NLP is the reason AI became accessible to non-programmers. When the interface to a system is ordinary language, the skill that matters is clear communication, not code. It is also why AI tools inherit the ambiguity of language: the same sentence can mean two things, and a model has to guess which one you meant.

Example at work

A customer service team routes thousands of inbound emails a week. An NLP pipeline classifies each by topic and urgency, extracts order numbers and account IDs, scores the customer's tone, and drafts a first reply for the agent to edit. The agent handles the judgment and the relationship.

Why it matters

Almost every AI feature aimed at office work is an NLP feature: summarizing, drafting, classifying, extracting, searching. Understanding that these all rest on the same underlying capability helps you see how one tool can do many jobs, and why clear inputs get better outputs.

Related terms