Evidence first.Judgement visible.
HAI turns research, vendor evidence and operational signals into workflow-first hospitality intelligence while keeping provenance, uncertainty and limitations visible.
The Verification Flow
Unlike traditional software directories, we process claims through a continuous five-stage lineage. Select any stage to inspect the methodology.
Source Process
We gather documentation, vendor assertions, and operator evidence where available to map the baseline boundary of product claims.
Published documentation and submitted evidence.
Map the stated operational claim.
What is the origin of the information?
This step highlights the progression of evidence evaluation. It shows how source material can move through review into caveated public intelligence without hiding uncertainty.
From source to public intelligence
HAI separates vendor claims, supporting material, assessment and public interpretation. We avoid generic numerical scores and keep material limitations visible.
Identify origin, scope and material limitations of the raw claims.
"We identify stated technical claims from available published documentation."
Record exactly what has and has not been reviewed.
Review vendor documentation. Review user evidence where available.
Evaluate evidence strength, operational context and uncertainty.
Document consistency: Consistent. Operational evidence: Caveated.
Apply judgement where context or consequence matters.
Human Review
Publish only what the evidence supports, with limitations visible.
Operational Integration may require documented manual steps to align workflow timing.
What readers can expect from every HAI review
Hospitality AI Index does not exist to sell software or write promotional copy. We compile independent intelligence to protect operator interests and clarify technological uncertainty.
Workflow-First
HAI organises intelligence around operational problems and workflows rather than technology categories alone. We evaluate software by how it integrates into real-world hospitality environments and day-to-day services.
Result: Operators receive clear reviews mapping technology directly to practical hospitality workflows.Evidence Separation
Vendor claims, vendor-submitted material and HAI analysis remain visibly distinguishable. We ensure that self-declared vendor statements never blend into verified, independent facts.
Result: Complete visual separation between direct claims and evaluated intelligence.Human Judgement
Automation can assist analysis, but consequential interpretation remains reviewable. We do not rely on automatic scoring or unmonitored machine metrics to determine findings.
Result: Thorough editorial oversight ensuring contextual accuracy.Visible Limitations
Weak, missing or conflicting evidence is shown rather than converted into false certainty. If key documentation is absent, we represent that gap transparently.
Result: Clarity on what is unknown, removing the theater of false certainty.Not every answer deserves the same level of confidence.
We categorise evidence using explicit states and visible caveats rather than a generic numerical score.
Verified Status
Available evidence reviewed by HAI supports the stated interpretation within the documented scope.
A VERIFIED status means the evidence reviewed by HAI is sufficient to support the stated interpretation within the stated scope and caveats.
- Evaluation Level
- VERIFIED EVIDENCE STATE
- Verification Focus
- Comprehensive Documentation
- Evidence Threshold
- Documentation and relevant submitted evidence reviewed by HAI.
- Potential Limitation
- Verified only for the documented environments, workflows, and evidence available for review.
Caveated Status
Useful evidence exists, but material limitations remain visible.
Available evidence supports limited utility, but operators must still consider visible caveats, manual workflows, or external dependencies.
- Evaluation Level
- CAVEATED EVIDENCE STATE
- Verification Focus
- Conditional Operational Utility
- Evidence Threshold
- Partial documentation or documented user evidence where available noting system limitations.
- Potential Limitation
- Manual checks, external dependencies, or workflow constraints may affect operations.
Insufficient Proof
HAI does not infer certainty where adequate evidence is absent.
If standard evidence or documented user evidence is absent, we do not imply endorsement or certainty.
- Evaluation Level
- UNVERIFIED EVIDENCE STATE
- Verification Focus
- Absent or Proprietary Data
- Evidence Threshold
- Documentation is missing, proprietary, or insufficient to support independent evaluation.
- Potential Limitation
- Available records are insufficient to support the technical assertions.
Automation can assist the analysis.Accountability stays human.
We do not treat automated summaries as self-validating, nor vendor claims as established fact. Claims with greater operational consequence receive additional human editorial review.
We collect official documentation, vendor assertions, and documented user evidence where available to build a traceable evidence record.
Human Accountability Covenants
We refuse to trust automated interpretations blindly. In the complex ecosystem of hospitality software and operational teams, unverified technology statements cause direct financial and service consequences.
Not procurement advice: HAI reviews evaluate evidence strength. They are not automatic procurement recommendations or endorsement products.
Not legal advice: We analyze functional integration capabilities and documented workflow claims; we do not provide legal or licensing assertions.
Commercial visibility does not buy editorial conclusions.
Editorial conclusions are intended to remain separate from commercial relationships and are presented with their evidence basis and limitations.
Editorial Intelligence
MERIT-ONLY & INDEPENDENTBrilliant 20 Inclusion Inclusion in the Brilliant 20 is an editorial decision. Commercial payment or sponsorship does not purchase inclusion.
No Editorial Purchase Hospitality AI Index does not sell rankings or custom endorsement badges. Commercial relationships do not purchase editorial findings or conclusions.
Independent Findings Evaluation states should reflect the evidence available to HAI at the time of review. Caveats and warnings remain visible while they continue to be supported by the record.
Commercial Visibility
TRANSPARENT BOUNDARIESSponsorship Distinctions Sponsored or commercial visibility on our platform is transparently declared and must always remain clearly distinguishable from editorial intelligence.
No Edit Influences Commercial or sponsorship relationships do not grant control over editorial findings or the right to remove evidence-based caveats.
Vendor-Submitted Material Material submitted directly by a vendor remains identified as vendor-submitted and visually distinct from HAI analysis and other evidence.
Evidence can change. Records do not change silently.
We maintain our index as an evolving record, but we do not modify active public states automatically based on requests. When new evidence or corrections are submitted, our editorial team manually reviews and assesses the material before changes are made.
Manual Review of Material: All submitted evidence and request details are reviewed manually through HAI editorial and research review to establish validity.
Caveat Visibility: Key limitations or corrections remain visible where appropriate in active public listings to safeguard operator confidence.
Submit evidence or corrections for editorial review.
HAI uses the governed Vendor Evidence route for methodology updates, correction requests and new evidence. This keeps the public submission path visible without creating a second submission surface.
Open Vendor Evidence