Unli.ai docs

Appearance

Managed Vector Store ​

Overview ​

Unli.ai provides a managed vector store that is available across all plans, offering seamless vector embedding and storage for your AI-powered projects.

Key Features ​

  • Universal Availability: Free and paid plans
  • Automatic Indexing: Seamless document processing
  • Built-in Embeddings: No additional configuration required
  • Scalable Storage: Optimized for AI document retrieval
  • Background Processing: Handled by Beesight system

Data Ingestion ​

The managed vector store automatically processes documents from:

  • Datasource Integrations
  • Direct Uploads
  • Web Scraping
  • API Imports

Metadata Management ​

Automatically stored metadata includes:

  • document_url
  • document_name
  • document_id
  • document_type
  • document_source
  • internal_id
  • document_body
  • document_page_number
  • document_total_pages
  • chunk_index

Custom Metadata Support ​

Additional metadata can be added via API payloads during document ingestion.

On paid plans, you can:

  • Bring Your Own Vector Store (BYOS)
  • Currently supported: Pinecone
  • Configure custom vector store settings

Performance and Limitations ​

  • Optimized for AI document retrieval
  • Suitable for most small to medium-sized document collections
  • Automatic scaling based on project needs

Getting Started ​

  1. Upload documents through Unli.ai interfaces
  2. Automatic vector store indexing begins immediately
  3. No additional configuration required

Support and Guidance ​

For specific questions about vector store capabilities, contact Unli.ai support.