Concepts
Understanding the core concepts of Endee Vector Database.
Projects
Projects are the highest level of hierarchy in Endee. A project is a store of indexes that organizes and isolates your vector data.
- Users can be members of multiple projects
- Each project contains one or more indexes
- Each project has its own set of users
Indexes
Indexes are the actual vector stores where your data lives. Endee uses dense indexes for storing and searching vectors.
- Each project can contain multiple indexes
- Dense vectors enable semantic search (also known as similarity search) — finding items based on meaning rather than exact keyword matches
- Indexes are optimized for fast Approximate Nearest Neighbor (ANN) searches
Vectors
Vectors are the fundamental data units stored in an index. Each vector represents your data as a numerical array (embedding) generated by an embedding model.
- Every vector is stored with a unique ID for identification and retrieval
- Vectors can have associated metadata for additional context
- Vectors can have filters for precise filtering during search queries
Auth Tokens
Auth tokens are used to authenticate your SDK requests with the Endee API.
- Generate auth tokens from your project dashboard
- Keep your tokens secure and never expose them in client-side code
- Tokens grant access to all indexes within a project
User Roles
Each project can have multiple users with different access levels:
| Role | Description |
|---|---|
| Admin | Full access to the project, can manage users, indexes, and settings |
| User | Limited access for standard operations like querying and upserting vectors |
Admins can invite new users and assign roles to control access within the project.