nuclia/nucliadb

NucliaDB, The AI Search database for RAG

What it solves

NucliaDB provides a way to store and search unstructured data using a hybrid approach. It eliminates the need for developers to manually handle data extraction, enrichment, and inference when paired with the Nuclia cloud ecosystem, making it easier to implement powerful NLP capabilities in applications.

How it works

It functions as a hybrid search database that combines vector, full-text, and graph indexes. Built with Rust and Python, it uses PostgreSQL as its storage layer and supports S3-compatible, GCS, and Azure Blob Storage for blobs. It can index fields, paragraphs, and semantic sentences, and integrates with the Nuclia Understanding and Learning APIs for AI-driven data transformation and model training.

Who it’s for

It is designed for developers building applications that require advanced search capabilities over large, unstructured datasets, specifically those needing multi-tenant support and a low-code path to NLP integration.

Highlights

  • Hybrid Search: Combines semantic vector search with traditional keyword and fuzzy search.
  • Flexible Storage: Supports text, files, vectors, labels, and annotations across various cloud blob storage providers.
  • Data Portability: Allows exporting data in formats compatible with PyTorch and HuggingFace datasets.
  • Enterprise Ready: Includes role-based security, multi-tenant support, and distributed search capabilities.

Related

  • Project
    HelixDB/helix-db

    A graph-vector database built in Rust that consolidates relational, vector, and graph data into a single platform for AI memory and knowledge graphs.

  • Project
    infiniflow/infinity

    An AI-native database designed for LLM applications that provides high-speed hybrid search across dense and sparse embeddings, tensors, and full-text data.

  • Project
    orneryd/NornicDB

    A Neo4j-compatible graph database that integrates vector retrieval and historical truth tracking into a single engine for AI agents and Graph-RAG.

  • Project
    datalevin/datalevin

    A durable, high-performance Datalog database that combines relational, document, and vector storage with built-in AI capabilities like in-DB embeddings and text generation.

  • Project
    manticoresoftware/manticoresearch

    A high-performance, open-source search database that serves as a fast, SQL-compatible alternative to Elasticsearch, supporting full-text, vector, and hybrid search.