Retrieval-augmented generation (RAG) is a technique that grounds an AI model's responses in trusted external data retrieved at query time. Instead of relying only on what a model learned in training, a RAG system fetches relevant documents — policies, products, records — and conditions the answer on them, improving accuracy and reducing hallucination.
Also known as: RAG, grounding
Enterprise answers must reflect the organization's own current, authoritative content. RAG keeps responses tied to source material you control and can cite, which is essential for trust and auditability.
We ground agents on each customer's policies, product data, and records so outputs are accurate, current, and traceable to a source.
A large AI model trained on broad data that can be adapted to many tasks through prompting or fine-tuning. Granite is IBM's enterprise family.
Read definition →Agentic AIAI systems that take action — applying logic, making decisions, and running multi-step workflows end to end across enterprise systems, with human oversight.
Read definition →GovernanceThe policies, controls, and processes that keep AI systems safe, compliant, transparent, and accountable across their lifecycle.
Read definition →