AI-Native

// How the category uses it

Used by the upstarts to imply structural superiority over legacy BI and POS platforms. The implication is that the AI is native to the architecture, not a feature bolted on top.

In practice, AI-native in this category means the platform was built using modern data infrastructure (data warehouses, embeddings, frontier model APIs) and ships an AI-generated narrative layer over the output. The underlying architecture — data ingestion, processing, dashboarding — is the modern data stack that any technology company uses.

// How superGM defines it

An AI-native architecture for restaurants would be designed around the problem AI is uniquely positioned to solve — detecting behavioral signals in a live environment and acting on them in sub-minute windows. Most platforms marketed as AI-native do not operate in live environments or sub-minute windows. The AI is the narrative. The architecture is the data stack.

// Why it matters

AI-native is used as a defensibility claim. In most deployments it is not defensible. Any competitor with a modern data stack can add a conversational AI layer over BI output. The moat in this category is either behavioral training data you own or an execution layer you built. The "native" modifier in front of AI is not either of those things.