Retrieval-Augmented Generation (RAG)

Retrieval-augmented generation (RAG) occurs when an artificial intelligence (AI) solution connects with external knowledge sources to provide more accurate and relevant results.

What Is Retrieval-Augmented Generation (RAG)?

Retrieval-augmented generation (RAG) is a process of enhancing the relevance and accuracy of an AI’s outputs. AI solutions are generally efficient at reviewing massive data sets to provide high-level learnings — much like how the human brain takes in its surroundings to make sense of the task in front of them. To complete more nuanced or in-depth tasks, however, both humans and AI need the right information and skills. Retrieval-augmented generation (RAG) enables an AI to link with external resources with more specific knowledge bases to improve the accuracy of their outputs. If a developer connects an AI assistant to market data about a specific product, for example, the AI could then draw from real-time data to provide a brand with more specific and relevant competitive insights. RAG can also include connecting with internal manuals, enhanced content (such as images or videos), and product descriptions at scale.

See Salsify in Action

Bring your team, tools, and data together to launch high-impact inbound marketing campaigns in less time.

Retrieval-Augmented Generation (RAG)

What Is Retrieval-Augmented Generation (RAG)?

Related Terms

Related Resources

The AI Revolution in Retail: Generative Engine Optimization for Customer Conversion

3 Ways Generative AI Augments Digital Commerce

See Salsify in Action