Methodology

Every material capability claim must be traceable to a dated public source and scoped to the relevant product family, edition, and deployment model.

Compared categories

Enterprise low-code platforms provide application models, integration, lifecycle, and runtime capabilities.
AI full-stack app builders generate and host applications with varying managed backend and code-portability models.
AI coding tools and agents modify repositories but normally leave the production stack to the engineering team.

Cross-category analysis compares responsibility boundaries rather than claiming that these products are equivalent.

Missing documentation is not evidence that a feature is absent.
“Open” does not automatically mean “open source.”
A data gateway is not an on-premises application runtime.
“No vendor lock-in” is replaced by concrete dependency and exit factors.
Prices, certifications, limits, and contractual rights are treated as volatile.
Facts are separated from editorial purchasing implications.
AI authoring, AI embedded in an application, external agent control, and the production runtime are evaluated separately.

The complete methodology, evidence states, review cadence, and limitations are maintained in METHODOLOGY.md.