This page defines the core concepts in Vectorify and how they relate to each other.
Chats in Vectorify are a conversational feature that allows you to have interactive discussions with your data. Chats enable you to start conversations and ask follow-up questions about your upserted data, creating a more natural and interactive way to query information than single queries.
The chat feature transforms your static data into an interactive knowledge base where you can have natural conversations and get contextual answers based on your upserted content.
An upsert in Vectorify is a robust operation that allows you to create or update items in a single operation. The term “upsert” combines “update” and “insert”, meaning you don’t have to worry about whether an item already exists.
Upserts are essential for keeping your data current in Vectorify, whether adding new content or updating existing information in your collections.
Collections in Vectorify are organizational containers that help you structure and manage your data effectively. Think of collections like folders that contain files or database tables that contain rows. They group related items together, making it easier to organize and retrieve your content.
Collections contain items that have been upserted into your Vectorify project. When you ask questions or query your data, collections help ensure you get the most relevant results by grouping contextually related content.
Items in Vectorify are the fundamental data units that contain your content and information. They are the individual pieces of content or data that you store in Vectorify. They’re organized within collections (which act like folders or database tables) and can be created or updated through upserts.
Items can include metadata — additional context information that helps you organize and filter your content. Metadata acts like tags or labels and provides essential filtering capabilities, though it’s not used directly for search.
Metadata in Vectorify is additional context information you can store alongside your items. Metadata stores extra information about your items that won’t be used for search but provides essential filtering capabilities. Think of it as tags or labels that help you organize and categorize your content.
You can include a metadata field when you create an upsert (which allows you to create or update items). This field stores whatever additional information you want to associate with that item, making it easier to organize and filter your data later.
Tenants are identifiers that specify which organization your data items belong to. They’re a way for different organizations or groups to separate and organize data within your Vectorify project, which is essential for SaaS projects.
When creating or updating items through upserts, you include a tenant field to specify which organization the items belong to. This ensures that data remains properly segmented and isolated between different tenants.