Skip to main content

​
Overview

Entity is a feature for managing named entities such as people, companies, products, and places mentioned in your project. Named entities in Knowledge, Episodes, and Topics are automatically detected and extracted, then centrally managed. This enables searches like “all information related to this person” or “all documents mentioning this company.”
Entities are extracted automatically. Simply register Knowledge or Topics, and the named entities they contain are automatically detected and registered as Entities via the Entities API.

​
What You Can Do with Entities

How Neuradex is Different: Manual vs AutomaticMost knowledge management tools require you to manually tag people and companies. Neuradex’s AI agent analyzes your content and automatically handles entity extraction, linking, and alias learning.

Person-Centric Search

Get “all information related to John” instantly. Search across meeting notes, documents, and conversations

Visualize Org Structure

Who is CEO of which company? Which company is a subsidiary of whom? Understand relationships at a glance

Handle Name Variations

“John Smith”, “Mr. Smith”, “John” all recognized as the same person. No more missed search results

Automatic Tagging

Just register documents, and people/companies mentioned are automatically tagged

​
Entity Types

TypeDescriptionExamples
personPeopleJohn Smith, 山田太郎
companyCompanies/OrganizationsGoogle, Acme Corp
productProducts/ServicesiPhone, Slack
placePlacesTokyo, Silicon Valley
otherOthersProject names, etc.

​
Aliases

When the same Entity has multiple names, you can register them as aliases. 
Entity: John Smith
├── Alias: John
├── Alias: Mr. Smith
└── Alias: J. Smith
By registering aliases, different spellings are recognized as the same Entity. Aliases are learned automatically and can also be added manually.

​
Entity Linking

Entities are linked to the following data types.

​
Linking with Knowledge

Entities in Knowledge content are automatically detected and linked. This enables:
  • Listing all Knowledge related to a specific Entity
  • Highlighting Entity mentions within Knowledge

​
Linking with Episodes & Topics

Entities appearing in conversations (Episodes) and Topics are similarly linked.

​
Entity Relations

Entities can have typed relationships. 13 relation types are available, allowing you to accurately model complex organizational structures.
Relation TypeDescriptionExample
CEO_OFIs CEO ofJohn Smith → Acme Corp
CTO_OFIs CTO ofJane Doe → Acme Corp
FOUNDER_OFFoundedElon Musk → SpaceX
WORKS_ATWorks atAlice → Google
SUBSIDIARY_OFIs subsidiary ofABC Labs → Acme Corp
PARTNER_OFPartner relationshipAcme Corp → DEF Inc
COMPETITOR_OFCompetes withSlack → Teams
INVESTOR_INInvests inVC Fund A → StartupXYZ
LOCATED_INLocated inAcme Corp → Tokyo
DEVELOPED_BYDeveloped byiPhone → Apple
ACQUIRED_BYAcquired byInstagram → Meta
SUPPLIES_TOSupplies toTSMC → Apple
RELATED_TOOther relationGeneral-purpose relation
These relations are automatically detected and suggested, and can also be manually configured via the Entities API. See the Entities Reference for the full data model.

​
Duplicate Detection and Merging

Similar Entities (e.g., “Acme Corp” and “Acme Corporation”) are automatically detected and can be merged. When merged, all links are automatically migrated.

​
Use Cases

Cross-search “all information about this person” across meeting notes, documents, and conversations.

​
Next Steps

Data Types

About Knowledge, Episodes, and Topics

Knowledge API

Knowledge API reference