Retrieval-Augmented Generation (RAG)

Retrieval Augmented Generation (RAG) is an architecture that augments the capabilities of a Large Language Model (LLM) like ChatGPT by adding an information retrieval system that provides grounding data. Adding an information retrieval system gives you control over grounding data used by an LLM when it formulates a response. For an enterprise solution, RAG architecture means that you can constrain generative AI to your enterprise content sourced from vectorized documents and images, and other data formats if you have embedding models for that content.

The decision about which information retrieval system to use is critical because it determines the inputs to the LLM. The information retrieval system should provide:

• Indexing strategies that load and refresh at scale, for all your content, at the frequency you require.
• Query capabilities and relevance tuning. The system should return relevant results, in the short form formats necessary for meeting the token length requirements of LLM inputs.
• Security, global reach, and reliability for both data and operations.
• Integration with embedding models for indexing, and chat models or language understanding models for retrieval.

As a Microsoft Partner we leverage Azure AI Search to build powerful RAG applications so when you create a search service with us, you work with the following capabilities:

• A search engine for vector search and full text hybrid search over a search index.
• Rich indexing with the ability to content transformation. This includes integrated data chunking and vectorization for RAG, lexical analysis for text, and optional applied AI for content extraction and enrichment.
• Rich query syntax for vector queries, text search, hybrid queries, fuzzy search, autocomplete, geo-search and others.
• Relevance and query performance tuning with semantic ranking, scoring profiles, quantization for vector queries, and parameters for controlling query behaviors at runtime.
• Azure scale, security, and reach.
• Azure integration at the data layer, machine learning layer, Azure AI services and Azure OpenAI.

At Digital Bricks we help you in setting up the right query and strategy to find relevant output within your datasets. Therefore, we leverage different query features.