What is the correct order of steps for building a RAG application using PL/SQL in Oracle Database 23ai?

What is the correct order of steps for building a RAG application using PL/SQL in Oracle Database 23ai?

• A. Load Document, Load ONNX Model, Create Embeddings, Perform Vector Search
• B. Load Document, Load ONNX Model, Split Text, Generate Output
• C. Load Document, Load ONNX Model, Split Text, Create Embeddings, Perform Vector Search
• D. Load Document, Load ONNX Model, Create Embeddings, Generate Output

Answer: (C)

Building a Retrieval-Augmented Generation (RAG) application using PL/SQL in Oracle Database 23ai involves several crucial steps that are executed in a specific sequence to ensure the application functions correctly. The correct sequence begins with loading the document, which is essential as it serves as the source material from which the application will draw information. Following that, loading the ONNX model is vital; this model enables the application to perform machine learning tasks by leveraging pre-trained models for efficient inference. Once the document and the model are in place, the next step is to split the text. This process involves breaking down the loaded document into manageable segments that can be processed more easily. This is an important step for optimizing the creation of embeddings, as smaller chunks of text can lead to more accurate representations of the information. After the text is split, the creation of embeddings is performed. This stage is crucial as embeddings transform the textual data into a vector format that can be used for similarity searches. With embeddings generated, the application can then proceed to perform vector searches, allowing it to retrieve relevant information based on user queries. This logical progression—loading the document, loading the model, splitting text, creating embeddings, and finally performing vector searches—ensures that the RAG

Building a Retrieval-Augmented Generation (RAG) application using PL/SQL in Oracle Database 23ai involves several crucial steps that are executed in a specific sequence to ensure the application functions correctly.

The correct sequence begins with loading the document, which is essential as it serves as the source material from which the application will draw information. Following that, loading the ONNX model is vital; this model enables the application to perform machine learning tasks by leveraging pre-trained models for efficient inference.

Once the document and the model are in place, the next step is to split the text. This process involves breaking down the loaded document into manageable segments that can be processed more easily. This is an important step for optimizing the creation of embeddings, as smaller chunks of text can lead to more accurate representations of the information.

After the text is split, the creation of embeddings is performed. This stage is crucial as embeddings transform the textual data into a vector format that can be used for similarity searches. With embeddings generated, the application can then proceed to perform vector searches, allowing it to retrieve relevant information based on user queries.

This logical progression—loading the document, loading the model, splitting text, creating embeddings, and finally performing vector searches—ensures that the RAG