AI Search and Now platform

[dheerajcheepati]

To answer the question of how AI Search works on the Now platform, we need to understand two important concepts on what constitutes AI Search and what is RAG.

What is AI Search?

AI search refers to the use of artificial intelligence technologies to enhance search capabilities. This can include a variety of applications, such as:

1. Semantic Search: AI can understand the context and meaning behind search queries, allowing it to return results that are more relevant to the user's intent, even if the exact keywords are not used.

2. Personalization: AI algorithms can learn from a user's past behavior to personalize search results, making them more relevant to the individual's interests and preferences.

3. Natural Language Processing (NLP): AI can process and understand human language, enabling users to make search queries in natural, conversational language.

4. Image and Voice Search: AI can analyze images and voice data to conduct searches based on visual content or spoken queries.

5. Predictive Search: AI can predict what a user is likely to search for based on current trends, user behavior, and other factors, sometimes offering suggestions before the user even completes typing their query.

6. Automated Indexing and Classification: AI can automatically categorize and index content, making it easier to retrieve relevant information from large datasets.

7. Machine Learning: Over time, AI search systems can learn and improve their accuracy and efficiency, providing better search experiences for users.

What is RAG?

In the context of artificial intelligence, RAG stands for "Retrieval-Augmented Generation." It is a technique used in natural language processing (NLP) that combines the power of language models with information retrieval.

Here is how RAG works -

1. Retrieval: When a question or prompt is given to the system, the RAG model first retrieves relevant documents or passages from a large corpus of text.
2. Augmentation: The retrieved documents are then provided as additional context to a language generation model.
3. Generation: The language model, which has been trained on vast amounts of text data, generates a response.

The RAG approach allows AI models to produce responses that are informed by specific information from external sources, rather than relying solely on the knowledge that was encoded into the model during its training.

How does AI Search Work on the Now Platform?

The embedded article clearly summarizes the inner workings of Now Assist in AI Search that spans an 8 steps approach at high-level.

I also found the Semantic vector search in AI Search very useful in understanding the concept of Hybrid Search which includes Keyword/Semantic search.