Astra Vectorize

Astra Vectorize lets the Astra DB database itself calculate the numeric fingerprint of your documents’ meaning, instead of an embeddings component. That fingerprint is what allows NappAI to find documents that say something similar to what you are looking for, even when they use different words, for example in a knowledge base.

The component does not calculate anything itself: it only tells Astra DB which provider and model to use. The provider must be set up in your Astra DB database.

When to use it

• You already keep your knowledge base in Astra DB and prefer the database itself to calculate the fingerprints.
• Your Astra DB collection already has an embeddings provider set up and you want to use it from your flows.
• You want to use a provider that Astra DB offers, such as Jina AI, Upstage or Voyage AI.

Configuration

Field	What it does
Credential (required)	The credential of the service DATASTAX Api. Pick one with Choose an option or create one with Add new credential. See Credentials.
Provider (required)	The provider Astra DB uses to calculate the fingerprints: Azure OpenAI, Hugging Face - Dedicated, Hugging Face - Serverless, Jina AI, Mistral AI, NVIDIA, OpenAI, Upstage or Voyage AI.
Model Name (required)	The model of the chosen provider. See the table of models below.

Provider	Models
Azure OpenAI	text-embedding-3-small, text-embedding-3-large, text-embedding-ada-002
Hugging Face - Dedicated	endpoint-defined-model
Hugging Face - Serverless	sentence-transformers/all-MiniLM-L6-v2, intfloat/multilingual-e5-large, intfloat/multilingual-e5-large-instruct, BAAI/bge-small-en-v1.5, BAAI/bge-base-en-v1.5, BAAI/bge-large-en-v1.5
Jina AI	jina-embeddings-v2-base-en, jina-embeddings-v2-base-de, jina-embeddings-v2-base-es, jina-embeddings-v2-base-code, jina-embeddings-v2-base-zh
Mistral AI	mistral-embed
NVIDIA	NV-Embed-QA
OpenAI	text-embedding-3-small, text-embedding-3-large, text-embedding-ada-002
Upstage	solar-embedding-1-large
Voyage AI	voyage-large-2-instruct, voyage-law-2, voyage-code-2, voyage-large-2, voyage-2

Advanced options

Field	What it does
Model Parameters	Extra parameters the provider asks for in Astra DB, as name and value pairs. Leave it empty if your provider does not need any.

Output

• Vectorize: the provider and model settings. Connect it to the Embedding or Astra Vectorize input of Astra DB. For the other vector databases, use an embeddings component such as OpenAI Embeddings.

Use the same model to store and to search

Fingerprints from two different models cannot be compared with each other. Use the same provider and the same model to store the documents and to search them.

Example

An agricultural cooperative, Cooperativa Campo Verde, keeps its crop sheets in Astra DB.

1. Add Astra Vectorize, choose the credential, set Provider to OpenAI and Model Name to text-embedding-3-small.
2. Connect its Vectorize output to the Embedding or Astra Vectorize input of Astra DB.
3. Astra DB calculates the fingerprints when it stores the sheets and when it searches them, and the cooperative’s assistant finds the sheet that answers each question.

Related components

• Astra DB
• OpenAI Embeddings