SAITAT / INFORMATION

Editorial methodology

How Saitat selects, classifies, compares and verifies software information.

Editorial principles

We prioritize products developers can meaningfully compare or integrate into real projects. A listing is not an endorsement. Inclusion is separate from commercial participation.

Evidence and freshness

Each listing links to an official source when available. Vendor descriptions, licenses, plans and APIs can change. If a listing says Not yet verified, it has not undergone a documented manual check in Saitat's verification log. An unverified pricing model is an editorial starting point, not a live quote.

Saitat intentionally omits exact prices when they cannot be reliably maintained. It does not fabricate user ratings, usage numbers, benchmark results or vendor availability.

Comparisons and alternatives

Comparison articles explain an intended use case and trade-offs. Automated alternatives are based on shared categories and are not a guarantee of feature-level parity. A choice can be optimal for one team's constraints and inappropriate for another.

AI assistance

AI assists with catalogue enrichment, drafting architectures, reviewing technical claims, and identifying required source evidence. Some reference stacks may be published automatically when configured safety and evidence checks pass. Higher-risk findings, contradictory evidence and unqualified material claims are held for correction or editorial attention. Automated review is not independent certification or evidence that a deployment has been tested.

Stack evidence and qualification

Each reference stack distinguishes architectural review, retrieved official sources, and outstanding implementation prerequisites. A source retrieval date documents when a page was fetched; it does not establish that all of its claims, pricing or API behavior were verified. A qualified reference may include disclosed, unresolved deployment prerequisites and must not be described as production-ready. Actual executable examples and test outcomes are separately labeled, and test results apply only to the recorded artifact and environment.

Publication and corrections

Content and provider terms can change after publication. Readers should verify provider documentation and test their chosen versions. Substantive corrections can be made through editorial review or automated monitoring; the site does not promise continuous independent testing of every product or stack.

Commercial independence

Paid referrals must be visibly disclosed. Sponsored status does not automatically improve editorial rankings. The publisher also operates GreatHosts, a hosting business that may appear in relevant listings. Hosting recommendations should use the same documented criteria as other providers.