Module utils 

Utility functions and helpers for common semantic search workflows

This module provides convenience functions that combine multiple features and simplify common patterns in semantic search applications.

Structs

BatchDeletionResult

Result of a batch deletion operation

BatchEmbeddingStats

Statistics for a batch embedding operation

BatchIndexResult

Result of a batch index operation with statistics

HealthCheckResult

Health check result for a semantic router

IndexStats

Export index statistics to a JSON-serializable structure

Functions

average_embedding

Calculate the average embedding from a set of embeddings

batch_delete

Delete multiple CIDs from a vector index in batch

cosine_similarity

Calculate cosine similarity between two embeddings

create_hybrid_index_from_map

Create a hybrid index with metadata extracted from a CID mapping

export_index_stats

health_check

Perform a comprehensive health check on a vector index

index_with_quality_check

Validate and index embeddings in a single operation with quality checking

normalize_vector

Normalize a vector to unit length (L2 norm = 1)

normalize_vectors

Normalize a batch of vectors in place

pairwise_similarities

Calculate pairwise cosine similarities between a query and multiple embeddings

validate_embeddings

Validate a batch of embeddings and return quality reports