High-Level Database
A high-level database like OneNode DB unifies NoSQL, vector, and object storage, simplifying data management for LLM applications. Developers can handle diverse data types without managing multiple databases or pipelines—OneNode DB takes care of it all in one streamlined system.
Low-Level Database
A low-level database requires developers to manage individual database types (e.g., NoSQL, vector, object storage) and build the necessary pipelines to connect them. This approach demands specialized knowledge and manual setup, making it more complex and time-consuming for handling diverse data types in LLM applications.
Database development is often complex and time-consuming
Choosing a Database Type
Setting Up the Database Server
Designing the Data Model
Building the Data Pipelines
Writing Queries and Operations
Scaling the Database
OneNode DB: A Single Database That Handles It All
With OneNode DB, all the complexities of managing multiple databases are eliminated.
Much like how Python and JavaScript simplify programming compared to low-level languages, OneNode DB unifies the full capabilities of NoSQL, vector databases, and object storageinto a single system.
It handles everything from data storage and indexing to integration and scaling, allowing developers to focus on building their applications without the hassle of managing separate databases or setting up complex pipelines.
{
"_id": { "$oid": "64d2f8f01234abcd5678ef90" },
"name": "Alice",
"profile": {
"bio": {
"@embText": {
"text": "Alice is a data scientist with expertise in AI and machine learning.",
"emb_model": "gpt-4o-mini"
}
},
"profile_picture": {
"@embImage": {
"data": "<binary image data>",
"mime_type": "image/jpeg",
"vision_model": "gpt-4o-mini"
}
}
}
}
OneNode Extended JSON
OneNode Extended JSON is designed to simplify AI-driven applications by embedding and indexing complex data types like text and images directly in the database. With no need for external pipelines, it handles everything from storage to semantic search automatically, saving developers time and effort. This powerful format makes managing both structured and unstructured data seamless, ideal for LLM applications.
Why OneNode Extended JSON is So Advanced
Automated Data Processing
OneNode Extended JSON automates complex tasks like image-to-text conversion, text chunking, embedding generation, and vector indexing, saving developers from manually handling these processes.
Customizable Embeddings
Developers can choose the embedding model for each JSON field, giving flexibility and control over how text and images are processed for semantic search.
Seamless Nested Fields
OneNode Extended JSON can be placed in any nested field, allowing complex data structures to be stored and indexed efficiently without constraints.
Effortless Media Storage
Store large media files, such as images, directly within the database without needing to deal with separate object storage systems.
Signed URLs for Media
Every saved media file is automatically assigned a signed URL, making it easy to retrieve and securely share stored media.
Customizable Parameters
Easily adjust parameters like chunk size and embedding models to fine-tune how data is processed and indexed.
Asynchronous Processing
OneNode processes data asynchronously, ensuring your users don’t have to wait while embedding and indexing happens in the background.
MongoDB Compatibility
Fully MongoDB-compatible, OneNode Extended JSON benefits from MongoDB’s rich ecosystem, extensive documentation, and community support.