Skip to content
Nappai Documentation

AI/ML Embeddings

This component is a key part of Nappai’s ability to understand and work with your data. It transforms your text or other data into numerical codes that computers can easily process. Think of it as translating your data into a language that Nappai’s AI understands. This translation is crucial for tasks like searching, comparing, and analyzing information.

Relationship with AI/ML API

This component uses a powerful AI/ML API to generate the embeddings. This API is a sophisticated service that performs the complex mathematical calculations needed to create these numerical representations. You don’t need to worry about the technical details; this component handles all the interaction with the API for you.

Inputs

  • Model Name: Choose the AI model Nappai will use to create the embeddings. Different models are better suited for different types of data and tasks. You’ll select from a list of pre-defined models. This is a required field.
  • Credential: This is an identifier that allows Nappai to securely access the AI/ML API. You will typically obtain this from your Nappai administrator.

Outputs

This component produces a set of embeddings. These are numerical values that represent your input data. These embeddings are then used by other Nappai components, such as vector databases, to perform powerful searches and analyses. You won’t directly interact with the embeddings themselves; instead, you’ll see the results of their use in other parts of the dashboard.

Usage Example

Imagine you have a large collection of customer reviews. You can use this component to create embeddings for each review. Then, you can use a vector database (like the ones listed below) to quickly find reviews that are similar to a specific query. This allows you to easily identify trends, positive and negative feedback, and much more.

Templates

[List of templates where the component can be seen and its configuration - This section requires information not provided in the prompt.]

  • Semantic Text Splitter: This component breaks down large texts into smaller, meaningful chunks before creating embeddings, improving accuracy.
  • Couchbase, Upstash, Chroma DB, Weaviate, Vectara, Redis, PGVector, FAISS, Astra DB, Qdrant, Pinecone, MongoDB Atlas, Milvus, Supabase, Cassandra, Text Embedder: These are vector databases that store and search the embeddings generated by this component, enabling powerful search and analysis capabilities.

Tips and Best Practices

  • Choose the appropriate model based on your data type and the task you are performing. Experiment with different models to find the best one for your needs.
  • Ensure your credential is correctly configured. Contact your Nappai administrator if you have any issues.

Security Considerations

The credential input is crucial for secure access to the AI/ML API. Never share your credential with unauthorized individuals. Report any suspicious activity to your Nappai administrator immediately.