Foundation Models

A foundation model is a large AI model trained on broad, general-purpose data that can be adapted to many downstream tasks rather than built for one. Large language models are the most common type. Enterprise software typically uses foundation models as the reasoning engine and adds its own knowledge sources, tools, guardrails, and governance around them.

Foundation models come as closed-source commercial models accessed through cloud providers, or open-source models that can be fine-tuned and deployed privately. Mature platforms route between several based on task type, latency, sensitivity, and cost through a gateway layer rather than depending on one. The misconception is that the foundation model is the product. It is not; the product is what is built around it, which is where domain knowledge, permissions, and auditability live.