askexperts/dist/rag/interfaces.d.ts

AskExperts SDK: build and use AI experts - ask them questions and pay with bitcoin on an open protocol

/**
 * Represents a chunk of text with its position information and embedding.
 */
export interface Chunk {
    /** Index of the chunk */
    index: number;
    /** Offset in the original text */
    offset: number;
    /** The text content of the chunk */
    text: string;
    /** The embedding vector (optional, added after embedding) */
    embedding: number[];
}
/**
 * Interface for text embedding services.
 * Responsible for splitting text into chunks and generating embeddings.
 */
export interface RagEmbeddings {
    /**
     * Initializes the embedding model.
     * Must be called before using the embed method.
     * @throws Error if already initialized
     * @returns Promise that resolves when initialization is complete
     */
    start(): Promise<void>;
    /**
     * Embeds the given text by splitting it into chunks and generating embeddings.
     * @param text The text to embed
     * @returns Promise resolving to an array of chunks with their embeddings
     * @throws Error if start() has not been called
     */
    embed(text: string): Promise<Chunk[]>;
}
/**
 * Result of a RAG search operation.
 */
export interface RagResult {
    /** Unique identifier for the vector */
    id: string;
    /** The embedding vector */
    vector: number[];
    /** Additional metadata associated with the vector */
    metadata: any;
    /** Distance/similarity score between the query vector and this result */
    distance: number;
}
/**
 * Document to be stored in the RAG database
 */
export interface RagDocument {
    /** Unique identifier for the vector */
    id: string;
    /** The embedding vector */
    vector: number[];
    /** Additional metadata to store with the vector */
    metadata: any;
}
/**
 * Interface for RAG database operations.
 * Provides methods for storing and searching vector embeddings.
 */
export interface RagDB {
    /**
     * Stores a single vector embedding with associated metadata in the specified collection.
     * @param collectionName The name of the collection to store the vector in
     * @param id Unique identifier for the vector
     * @param vector The embedding vector (array of numbers)
     * @param metadata Additional metadata to store with the vector
     */
    store(collectionName: string, id: string, vector: number[], metadata: any): Promise<void>;
    /**
     * Stores multiple vector embeddings with associated metadata in the specified collection.
     * @param collectionName The name of the collection to store the vectors in
     * @param documents Array of documents to store
     */
    storeBatch(collectionName: string, documents: RagDocument[]): Promise<void>;
    /**
     * Searches for similar vectors in the specified collection.
     * @param collectionName The name of the collection to search in
     * @param vector The query vector to search for
     * @param limit Maximum number of results to return
     * @returns Promise resolving to an array of RagResult objects
     */
    search(collectionName: string, vector: number[], limit: number): Promise<RagResult[]>;
    /**
     * Searches for similar vectors in the specified collection using multiple query vectors.
     * @param collectionName The name of the collection to search in
     * @param vectors Array of query vectors to search for
     * @param limit Maximum number of results to return per query vector
     * @returns Promise resolving to an array of arrays of RagResult objects, one array per query vector
     */
    searchBatch(collectionName: string, vectors: number[][], limit: number): Promise<RagResult[][]>;
}