# What is natural language processing (NLP)? > Natural language processing (NLP) is how computers read and work with everyday human language. Plain definition, a meeting-transcript example and mix-ups. Source: https://learn.ynteractive.com/content/glossary/natural-language-processing · Updated 2026-10-01 · Free from Stepthrough (https://learn.ynteractive.com) [AI basics](https://learn.ynteractive.com/content/glossary#topic-ai-basics) · [AI glossary](https://learn.ynteractive.com/content/glossary) Updated October 1, 2026 · By [Robert Breen](https://learn.ynteractive.com/content/about) Natural language processing (NLP) is the branch of AI that lets computers read, interpret and produce everyday human language, the way people actually write and speak. It covers tasks like summarizing, translating, pulling out names and dates, sorting messages by topic and answering questions. ## Why it matters for a small business A lot of business information lives in messy language: emails, call notes, meeting transcripts, reviews, support tickets. Spreadsheets and databases cannot do much with a paragraph. NLP is what turns that paragraph into something you can act on, like a list of decisions, a due date, a category or a reply. Older NLP tools needed a separate system for each job: one to detect sentiment, another to find dates. Large language models now handle most of these jobs from a plain-English request. That is why one well-written prompt can replace what used to be several specialized tools, but it also means you have to say exactly what to pull out. ## In a real lesson: Turn a Client Meeting into Action Items and a Follow-Up Email (Accounting Firms) The [client meeting recap lesson](https://learn.ynteractive.com/content/chatgpt-client-meeting-recap-accounting-firm) starts with exactly the kind of text NLP exists for: a Google Meet transcript of Millbrook Accounting's Q3 review with Cedar Lane Landscaping, both made-up businesses. Four people talk over each other, there are a few "ums", and the useful facts are buried in conversation: statements uploaded "by the twentieth", twenty-three uncategorized transactions, a truck bought August twelfth, three missing W-9s. You copy the transcript with **Edit**, **Select all** and **Copy**, then click **Paste request** in ChatGPT. The request asks for a three-sentence summary, the decisions, an action items table with task, owner and due date, and open questions, and ends with: "Only use what's in the transcript. If an owner or date isn't said, write `Not stated`." That is several classic NLP jobs in one message: summarizing, extracting who promised what, recognizing dates and spotting unresolved questions. The result is a table you can check, and two rows say Not stated because nobody said when the truck paperwork or the bookkeeping quote was due. [Try this lesson free](https://learn.ynteractive.com/modules/accounting-meeting-summary) or [read the step-by-step guide](https://learn.ynteractive.com/content/chatgpt-client-meeting-recap-accounting-firm). ## Common confusions ### NLP vs large language models NLP is the field: getting computers to work with human language. [Large language models](https://learn.ynteractive.com/content/glossary/large-language-model) are the current, most capable tool for it. Not all NLP uses an LLM, but most new business tools that read text do. ### NLP vs keyword matching A keyword filter only sees exact words, so "can you move my appointment" and "I need a different day" look unrelated. NLP works with meaning, so both are recognized as a reschedule request. ## Tips - Name the exact pieces you want pulled out (owner, due date, amount) instead of asking for "the important stuff". - Ask for a placeholder like Not stated when information is missing, so gaps are visible instead of guessed. - Check extracted dates and names against the source; transcripts often contain mishearings. ## Related terms [Large language model (LLM)](https://learn.ynteractive.com/content/glossary/large-language-model) · [AI summarization](https://learn.ynteractive.com/content/glossary/ai-summarization) · [Data extraction](https://learn.ynteractive.com/content/glossary/data-extraction) · [Meeting transcript](https://learn.ynteractive.com/content/glossary/meeting-transcript) · [Sentiment analysis](https://learn.ynteractive.com/content/glossary/sentiment-analysis) ## Where you use it: free lessons - [Turn a Client Meeting into Action Items and a Follow-Up Email (Accounting Firms)](https://learn.ynteractive.com/content/chatgpt-client-meeting-recap-accounting-firm) (ChatGPT, 9 min) - [Summarize a Meeting Transcript with ChatGPT](https://learn.ynteractive.com/content/chatgpt-summarize-meeting-transcript) (ChatGPT, 9 min) - [AI Email Responder: Draft Gmail Replies Automatically with n8n](https://learn.ynteractive.com/content/ai-email-responder-gmail-drafts-n8n) (n8n, 12 min) ## Frequently asked questions **Is ChatGPT an NLP tool?** Yes. Reading your message and writing a reply in plain language is natural language processing. ChatGPT handles summarizing, extracting and rewriting from one chat box. **Can NLP handle typos and casual writing?** Modern language models handle typos, slang and rambling speech well, which is why a raw meeting transcript works. Unusual names and numbers are the parts most worth double-checking. [All AI glossary terms, A to Z](https://learn.ynteractive.com/content/glossary) · [Free prompt templates](https://learn.ynteractive.com/content/prompts)