RAG Queries

Retrieval-Augmented Generation (RAG) Queries enrich the capabilities of Hyper by combining the retrieval of relevant data with the generation capabilities of language models. This approach enables the development of applications that can understand and generate nuanced responses based on the live data synced with Hyper.

Understanding RAG Queries

RAG Queries in Hyper leverage the power of both retrieval and generative AI to provide sophisticated answers to complex queries. By first identifying the most relevant information from connected data sources, and then processing this information through generative models, RAG Queries can produce contextually rich, accurate, and informative responses.

Available RAG Query Methods

Hyper supports several RAG Query methods, each designed for specific use cases:

1. Information Retrieval: Extracts relevant information from the dataset to answer specific queries directly.

   • Information Retrieval Guide

2. Question & Answer: Focuses on generating precise answers to questions, based on the information retrieved from the data sources.

   • Question & Answer Guide

3. Summarization: Creates concise summaries from larger texts, enabling quick understanding of lengthy documents without the need for manual review.

   • Summarization Guide

4. Semantic Search: Enhances search capabilities by understanding the semantic context of queries, returning results that are conceptually related to the search terms, not just textually similar.

   • Semantic Search Guide

5. Deep Insight: Generates in-depth analyses and insights from data, providing comprehensive answers that go beyond surface-level information.

   • Deep Insight Guide

Getting Started

To integrate RAG Queries into your application, begin by establishing connections to your data sources using Hyper's Data Connectors. Once your data is accessible, you can utilize the RAG Queries API to construct and execute queries that leverage the full spectrum of Hyper's retrieval and generation capabilities.