Named-entity recognition (NER) is the NLP task of locating and classifying named things in text, such as people, places, dates, products, and order numbers, so software can extract structured data from unstructured language.
When a customer writes "I want to return the blue jacket from order 48213 I bought last Tuesday," NER is what isolates blue jacket as a product, 48213 as an order number, and last Tuesday as a date. It turns a sentence into fields a system can use. NER is the core of entity extraction: entity extraction software applies it to pull structured fields like these out of free text.
NER is closely tied to slot filling: NER finds the entities in the text, and slot filling decides which of them satisfy the requirements of the recognized intent. One detects, the other assigns.
NER is a commodity capability, and Aide, the agentic AI platform for customer experience, uses it the way any serious system does. What matters is the order of operations. Aide classifies intent first, then uses entity extraction to populate the slots that intent requires, while the Customer Context Engine cross-checks extracted entities against real records in Shopify, WooCommerce, or Salesforce.
Validation is the safeguard. An extracted entity is treated as unverified until it matches a real record: if an order number does not correspond to an actual order, no automation proceeds on it. What was extracted, and what it was checked against, stays recorded and reviewable.