Skip to content
Nappai Documentation

MongoDB Atlas

The MongoDB Atlas component, known as MongoVectorStoreComponent, is designed to interact with MongoDB Atlas as a vector store. It provides vector search capabilities, allowing users to index and search documents within a MongoDB Atlas database.

Relationship with MongoDB Atlas

This component connects to MongoDB Atlas, a cloud-based database service, to store and retrieve data using vector search. It leverages MongoDB’s capabilities to manage and query large datasets efficiently.

Inputs

  • Database Name: The name of the database in MongoDB Atlas where your data is stored.
  • Collection Name: The specific collection within the database that contains your documents.
  • Index Name: The index used for vector search, which helps in efficiently finding documents.
  • Search Query: A query to search for specific documents in the vector store.
  • Ingest Data: A list of data items to be added to the vector store.
  • Embedding: The embedding used for vector search, which helps in finding similar documents.
  • Number of Results: The number of search results to return, with a default of 4.
  • Credential: The credential required to access MongoDB Atlas securely.

Outputs

The component produces a list of documents that match the vector search query. These documents are transformed into a format that can be easily used in workflows, such as further processing or analysis.

Usage Example

Imagine you have a collection of customer reviews stored in MongoDB Atlas. By using this component, you can perform a vector search to find reviews similar to a specific query, such as “positive feedback on product quality.” This helps in quickly identifying relevant documents for analysis or reporting.

Templates

Currently, there are no specific templates where this component is pre-configured. However, it can be integrated into custom workflows within the Nappai system.

  • NVIDIA Rerank: Enhances search results by reranking documents using the NVIDIA API.
  • Multi Query Retriever: Retrieves data using multiple queries for comprehensive results.
  • Ensemble Retriever: Combines results from multiple retrievers for improved accuracy.
  • Retrieval QA: Facilitates question-answering by querying sources from a retriever.

Tips and Best Practices

  • Ensure your MongoDB Atlas credentials are securely stored and managed.
  • Use descriptive names for your database, collection, and index to make it easier to manage and understand your data structure.
  • Regularly update your vector store with new data to keep your search results relevant.

Security Considerations

  • Always use secure credentials to access MongoDB Atlas.
  • Ensure that your data ingestion processes comply with data privacy regulations and best practices.