Completions

Keep-customer swarm · Review Response Agent · Review-classification skill · Build pillar · Published August 31, 2026

How to build multi-location review classification at scale

At 50-500 locations every connected platform generates 100-10,000 reviews per day. A regulatory complaint about a slip-and-fall buried in 2,000 4-star “great food” reviews loses 6-72 hours of response window before a human reads it. This guide explains how to architect the review-classification skill on the Review Response Agent end-to-end at multi-location customer-review-monitoring scale: per-portfolio per-banner per-location per-canonical-review-source-pointer + per-canonical-multi-class-classifier-spec + per-canonical-priority-tier-spec + per-canonical-crisis-escalation-spec + per-canonical-language-coverage-spec + per-canonical-vertical-overlay-spec + per-canonical-anti-review-bomb-detection-spec + per-canonical-multi-platform-deduplication-spec + per-canonical-compliance-gate-spec + per-canonical-audit-trail + per-portfolio-audit-trail.

Engage the review-response agentSee the 32-agent swarm

What you will build

  • Per-portfolio per-banner per-location per-canonical-review-source-pointer across 30+ surfaces — GBP + Yelp + Healthgrades + Zocdoc + OpenTable + Resy + Tripadvisor + Booking.com + Hotels.com + Expedia + Foursquare + Facebook + Apple Maps + Bing + Apple App Store + Google Play + Better Business Bureau + Trustpilot + G2 + Capterra + Indeed + Glassdoor + Apartments.com + RateMDs + Vitals + Avvo + LegalMatch + Martindale-Hubbell + Sitejabber + ResellerRatings + Influenster.
  • Per-canonical-multi-class-classifier-spec — per-sentiment (multi-class polarity VADER + RoBERTa + XLM-R + mBERT + LayoutXLM + DistilBERT + DeBERTa + multilingual-MiniLM + LaBSE + multilingual-sentence-transformer ensemble) + per-concern (23 classes: operational + product + service + price + staff + facility + cleanliness + wait-time + accuracy + delivery + billing + returns + accessibility + reservation + parking + noise + safety + equipment + temperature + portion-size + menu-availability + payment-method + reservation-system) + per-crisis-indicator (30 classes including regulatory complaint + health hazard + assault claim + discrimination claim + legal threat + media attention + child endangerment + elder abuse + harassment + theft + fraud + active-shooter + fire-incident + food-poisoning cluster + bedbug + cockroach + rat + mold + asbestos + lead + Legionella + norovirus + salmonella + E.coli + Campylobacter + listeria + Clostridium) + per-intent (13 classes including complaint + question + positive-testimonial + review-bomb attack + competitor sabotage + employee-grievance-misposted + spam + accidental + bot-generated + paid-influencer-undisclosed + rage-quit + misunderstanding) + per-fake-detection (per-pattern-burst-anomaly + per-IP-cluster + per-account-newness + per-language-anomaly + per-sentiment-vs-rating-inconsistency + per-text-overlap-cluster + per-GPT-detector + per-Originality.ai + per-Copyleaks + per-Crossplag + per-undisclosed-insider-detection per FTC 2024 final rule + per-paid-review + per-coordinated-inauthentic-behavior + per-Tripadvisor-red-badge-fraud-signal-cross-check).
  • Per-canonical-priority-tier-spec + per-canonical-crisis-escalation-spec — P0 crisis ≤1hr SLA + P1 active churn ≤24hr SLA + P2 standard ≤72hr SLA + P3 positive ≤7day SLA + P4 archive + per-SLA-breach-alert + per-escalation-on-no-acknowledge + per-crisis-tier P0 legal-counsel notification + media-relations notification + regulatory-counsel notification + executive notification + crisis-comms-team mobilization + state DPH + FDA MedWatch + CDC + OSHA + DOJ + EEOC + CFPB + FTC + state AG notification when applicable.
  • Per-canonical-language-coverage-spec + per-canonical-vertical-overlay-spec — 80+ languages via Whisper + NLLB-200 + Google Translate + DeepL + AWS Translate + Azure Translate ensemble + 13 verticals (healthcare HIPAA no-acknowledge-as-patient + legal no-client-acknowledge + financial FINRA Rule 3110 supervisor + cannabis state board 12-state + alcohol TABC/DISCUS tied-house + restaurant FDA food safety + automotive Magnuson-Moss + childcare ESEA + senior care Medicare Part C + veterinary AVMA + real estate Fair Housing + fitness + beauty state cosmetology board).
  • Per-canonical-anti-review-bomb-detection-spec + per-canonical-multi-platform-deduplication-spec — per-burst-rate-anomaly (Z-score + Bayesian changepoint PELT + CUSUM + EWMA + Hawkes process self-exciting) + per-cohort-newness (IP cluster + account-creation-date cluster + language cluster) + per-text-similarity-cluster (MinHash-LSH + SimHash + sentence-transformer cosine + near-duplicate detection) + per-coordinated-inauthentic-behavior + per-Tripadvisor-red-badge / per-Yelp-recommended-vs-non-recommended / per-GBP-suspicious / per-Trustpilot-flagging cross-check + per-paid-influencer-undisclosed per FTC Endorsement Guides 2024 + per-same-reviewer-cross-platform-via-identity-graph + per-cross-platform-reply-consistency.
  • Per-canonical-compliance-gate-spec — FTC fake-review rule 2024 (16 CFR Part 465; civil penalties up to $51,744 per violation) + HIPAA marketing authorization 164.508 (Manasa Health Center 2023 $30k + Elite Dental 2019 $10k + UMass Memorial 2019 $230k OCR enforcement precedent; cannot acknowledge reviewer as patient) + FINRA Rule 3110 social-media supervisor + FINRA 2210 + state bar advertising Rule 7.1-7.3 + state medical/dental/veterinary/cosmetology boards + cannabis state board 12-state matrix + alcohol TABC/CalABC/SLA/DISCUS tied-house + tobacco FDA prohibition + FDA DSHEA + FDA OPDP + FTC substantiation Pfizer 1972 + EU AI Act Article 50 AI-disclosure + EU AI Act Article 13/14/15 + Digital Services Act Article 30/26 + Digital Markets Act + Tennessee ELVIS Act 2024 (when AI-voice in audio response) + 11-state deepfake-law matrix + Massachusetts AG Copley Advertising 2017 (when geofence-triggered healthcare review response) + Illinois BIPA + Texas CUBI + Washington MHMDA + Connecticut CTDPA + Texas DPSA + Virginia CDPA + Colorado CPA + Utah CPA + CCPA/CPRA sensitive PI + CCPA right-to-opt-out-of-automated-decisionmaking + GDPR Article 22 + WCAG 2.2 AA + ARIA + EAA EN 301 549 + Section 508 + ADA Title III + NIST AI RMF + ISO 42001 + ISO 27001 + SOC 2 Type II + OPA/Cedar/Casbin/Cerbos/Oso policy-as-code.
  • Per-canonical-cross-skill-handoff + per-canonical-audit-trail — per-handoff-to-33-sibling-skills + per-per-review-canonical-audit-record + per-immutable-WORM-storage + per-7-year-IRS-tax-retention + per-7-year-FTC-substantiation-retention + per-7-year-HIPAA-medical-record-retention + per-7-year-state-bar-record-retention + per-6-year-SEC-record-retention + per-3-year-FINRA-record-retention + per-3-year-FINRA-Rule-3110-archive + per-3-year-Illinois-BIPA-biometric-retention.

Why per-vendor-Birdeye-account-flat-classifier breaks at multi-location customer-review-monitoring scale

Per-vendor-Birdeye-canonical-account-flat-classifier ships per-account per-flat-rule-or-keyword primitive. At 1-account-1-flat-classifier scale per-account per-flat-rule primitive is enough. At multi-location customer-review-monitoring scale per-canonical-review-source-pointer + per-canonical-multi-class-classifier-spec + per-canonical-priority-tier-spec + per-canonical-crisis-escalation-spec + per-canonical-language-coverage-spec + per-canonical-vertical-overlay-spec + per-canonical-anti-review-bomb-detection-spec + per-canonical-multi-platform-deduplication-spec + per-canonical-compliance-gate-spec + per-canonical-audit-trail. Per-vendor-Podium + Yotpo + Reputation.com + Trustpilot-canonical-account-flat-classifier ship per-vendor per-native account-flat-classifier primitives.

The FTC-fake-review-rule-2024 anchor is the operationally distinctive constraint. 16 CFR Part 465 became effective in October 2024 and prohibits fake/AI-generated reviews, undisclosed insider reviews, suppression of negative reviews, buying or selling reviews, and review hijacking. Civil penalties run up to $51,744 per violation under the inflation-adjusted FTC Act Section 5(m) authority. The classifier must surface AI-generated reviews (GPT-detector + Originality.ai + Copyleaks + Crossplag ensemble) and undisclosed-insider patterns, and the response side must not suppress legitimate negative reviews.

The HIPAA-no-acknowledge-as-patient anchor is the second distinctive constraint. Office for Civil Rights enforcement precedent — Manasa Health Center 2023 $30,000 settlement, 2022 NJ practice $30,000 settlement, Elite Dental Associates 2019 $10,000 settlement, UMass Memorial Medical Center 2019 $230,000 settlement — establishes that healthcare response cannot acknowledge reviewer as patient even if reviewer self-discloses, cannot disclose patient information, and cannot dispute factual claims that would require disclosing PHI.

The FINRA-Rule-3110-social-media-supervisor anchor is the third distinctive constraint. Financial-services review responses require pre-use principal supervisor approval per FINRA Regulatory Notice 10-06 + 11-39 + 17-18 social-media guidance. 3-year archive retention required per FINRA Rule 4511 + SEC Rule 17a-4. The classifier must route financial reviews to supervisor queue before drafting.

The Tennessee-ELVIS-Act-2024 anchor is the fourth distinctive constraint. When AI-voice synthesis renders an audio response to an audio review (TikTok/Instagram Reels/YouTube Shorts video reviews), the Tennessee Ensuring Likeness Voice and Image Security Act requires consent for voice cloning and creates a property right in voice identity. Eleven additional states (California + New York + Texas + Florida + Illinois + Indiana + Louisiana + Michigan + Minnesota + Tennessee + Washington) have parallel deepfake/voice statutes.

The Massachusetts-AG-Copley-Advertising-2017 anchor is the fifth distinctive constraint. When geofence-triggered healthcare review response touches healthcare facilities, the Massachusetts Attorney General consumer-protection precedent applies — Copley Advertising paid $50,000 to settle claims of geofencing women near Planned Parenthood clinics with anti-abortion ads. Geofence-based crisis response near healthcare facilities carries state-AG consumer-protection risk.

The operator-side architecture above per-vendor-flat-classifier primitive is canonical-review-source-pointer + multi-class-classifier-spec + priority-tier-spec + crisis-escalation-spec + language-coverage-spec + vertical-overlay-spec + anti-review-bomb-detection-spec + multi-platform-deduplication-spec + compliance-gate-spec + cross-skill-handoff + audit-trail + portfolio-audit-trail.

What is in market today

Per-platform per-review-management-vendor

Birdeye, Podium, Yotpo, Reputation.com, Trustpilot, Grade.us, ReviewTrackers, ReviewInc, NiceJob, GatherUp, BirdEye, Five Star Review System, Reputation Defender, Searchbloom, NetReputation, BrandYourself, Reputation Sciences. Per-account per-flat-rule-or-keyword primitive routes anything matching flat list to flat editorial queue. Per-canonical-review-source-pointer-canonical-multi-class-classifier-canonical-priority-tier-canonical-crisis-escalation-canonical-language-coverage-canonical-vertical-overlay-canonical-anti-review-bomb-detection-canonical-multi-platform-deduplication-canonical-compliance-gate-canonical-audit-trail is not the primitive.

Per-platform per-sentiment-classifier-vendor

AWS Comprehend, Azure Text Analytics, Google Cloud Natural Language, MonkeyLearn, Lexalytics, Repustate, Brandwatch, Talkwalker, Sprinklr, Khoros, Hootsuite Insights, Meltwater, Mention, Awario, Brand24, Cision, Critical Mention, Notified, Onclusive. Per-account per-flat-sentiment-score primitive (typically blind to per-concern 23-class + per-crisis-indicator 30-class + per-intent 13-class + per-fake-detection per FTC 2024 + per-Tripadvisor-red-badge-cross-check semantics). Per-canonical-per-review-multi-class-classifier-canonical-per-review-23-concern-class-canonical-per-review-30-crisis-indicator-class-canonical-per-review-13-intent-class-canonical-per-review-fake-detection-per-FTC-2024-canonical-per-review-Tripadvisor-red-badge-cross-check-canonical-per-review-classifier-confidence-tier-canonical-per-review-classifier-explainability is not the primitive.

Per-platform per-crisis-management-vendor

Edelman, Weber Shandwick, Ketchum, FleishmanHillard, Brunswick Group, Sard Verbinnen, Kekst CNC, Joele Frank, Teneo, Finsbury Glover Hering, FGS Global, Sitrick & Company, Levick, Stratacomm, Crisis Communications LLC, Bernstein Crisis Management, Burson Cohn & Wolfe. Per-account per-flat-crisis-playbook primitive (typically blind to per-canonical-crisis-tier P0/P1/P2/P3/P4 + per-canonical-crisis-escalation legal-counsel/media-relations/regulatory-counsel/executive/ crisis-comms-team/state-DPH/FDA-MedWatch/CDC/OSHA/DOJ/EEOC/ CFPB/FTC/state-AG-notification semantics). Per-canonical-per-crisis-tier-P0-legal-counsel-notification-canonical-per-crisis-tier-P0-media-relations-notification-canonical-per-crisis-tier-P0-regulatory-counsel-notification-canonical-per-crisis-tier-P0-executive-notification-canonical-per-crisis-tier-P0-crisis-comms-team-mobilization-canonical-per-crisis-tier-P0-state-DPH-FDA-MedWatch-CDC-OSHA-DOJ-EEOC-CFPB-FTC-state-AG-notification is not the primitive.

Per-platform per-vertical-compliance-vendor

Compliance.ai, Smarsh, Global Relay, Hearsay Systems, Erado, Actiance, Theta Lake, Behavox, Shield FC, Steel Eye, NICE Actimize, Cornerstone Compliance, OneTrust, TrustArc, Ketch, Securiti, BigID, DataGrail, Transcend, Osano. Per-account per-flat-vertical-rule primitive (typically blind to per-review FTC fake-review 2024 + HIPAA no-acknowledge-as-patient OCR enforcement precedent + FINRA Rule 3110 social-media supervisor + state bar advertising Rule 7.1-7.3 + state medical/dental/veterinary/cosmetology boards + cannabis state board 12-state + alcohol DISCUS tied-house + Massachusetts AG Copley Advertising 2017 + Tennessee ELVIS Act + 11-state deepfake-law matrix + Connecticut CTDPA + Texas DPSA + Virginia CDPA + Colorado CPA + Utah CPA semantics). Per-canonical-per-review-FTC-fake-review-2024-canonical-per-review-HIPAA-no-acknowledge-as-patient-canonical-per-review-FINRA-Rule-3110-canonical-per-review-state-bar-canonical-per-review-state-professional-licensing-canonical-per-review-cannabis-state-board-12-state-canonical-per-review-alcohol-DISCUS-tied-house-canonical-per-review-Massachusetts-AG-Copley-Advertising-2017-canonical-per-review-Tennessee-ELVIS-Act-canonical-per-review-11-state-deepfake-law-canonical-per-review-5-state-US-privacy-law is not the primitive.

How the architecture is built

  1. Per-portfolio per-banner per-location per-canonical-review-source-pointer-substrate. Per-30+-canonical-review-source canonical-source.
  2. Per-portfolio per-canonical-multi-class-classifier-spec. Per-sentiment-multi-class-polarity-ensemble + per-concern-23-class + per-crisis-indicator-30-class + per-intent-13-class + per-fake-detection-per-FTC-2024 + per-classifier-confidence-tier canonical-classifier.
  3. Per-portfolio per-canonical-priority-tier-spec. Per-P0-crisis-1hr-SLA + per-P1-active-churn-24hr-SLA + per-P2-standard-72hr-SLA + per-P3-positive-7day-SLA + per-P4-archive + per-tier-SLA-breach-alert + per-tier-escalation-on-no-acknowledge canonical-priority-tier.
  4. Per-portfolio per-canonical-crisis-escalation-spec. Per-legal-counsel + per-media-relations + per-regulatory-counsel + per-executive + per-crisis-comms-team + per-state-DPH + per-FDA-MedWatch + per-CDC + per-OSHA + per-DOJ + per-EEOC + per-CFPB + per-FTC + per-state-AG-notification + per-crisis-confidence-tier + per-crisis-FBC-feedback-loop canonical-crisis-escalation.
  5. Per-portfolio per-canonical-language-coverage-spec. Per-Whisper + per-NLLB-200 + per-Google-Translate + per-DeepL + per-AWS-Translate + per-Azure-Translate-80+-language-ensemble canonical-language-coverage.
  6. Per-portfolio per-canonical-vertical-overlay-spec. Per-healthcare-HIPAA + per-legal-state-bar + per-financial-FINRA + per-cannabis-state-board-12-state + per-alcohol-DISCUS + per-restaurant-FDA + per-automotive-Magnuson-Moss + per-childcare-ESEA + per-senior-care-Medicare-Part-C + per-veterinary-AVMA + per-realestate-Fair-Housing + per-fitness + per-beauty-cosmetology canonical-vertical-overlay.
  7. Per-portfolio per-canonical-anti-review-bomb-detection-spec. Per-burst-rate-Hawkes-process + per-cohort-newness-anomaly + per-text-similarity-MinHash-LSH-SimHash-sentence-transformer-cosine + per-coordinated-inauthentic-behavior + per-Tripadvisor-red-badge-Yelp-recommended-GBP-suspicious-Trustpilot-flagging-cross-check + per-paid-influencer-undisclosed-per-FTC-Endorsement-Guides-2024 + per-anti-review-bomb-confidence-tier canonical-anti-review-bomb-detection.
  8. Per-portfolio per-canonical-multi-platform-deduplication-spec. Per-same-reviewer-cross-platform-via-identity-graph + per-cross-platform-reply-consistency canonical-multi-platform-deduplication.
  9. Per-portfolio per-canonical-compliance-gate-spec. Per-FTC-fake-review-rule-2024 + per-HIPAA-marketing-authorization-164.508-no-acknowledge + per-FINRA-Rule-3110 + per-FINRA-2210 + per-state-bar-advertising + per-state-professional-licensing + per-cannabis-state-board-12-state + per-alcohol-TABC-CalABC-SLA-DISCUS-tied-house + per-tobacco-FDA-prohibition + per-FDA-DSHEA + per-FDA-OPDP + per-FTC-substantiation-Pfizer-1972 + per-EU-AI-Act-Article-50-AI-disclosure + per-EU-AI-Act-Article-13-14-15 + per-Digital-Services-Act-Article-30-Article-26 + per-Digital-Markets-Act + per-Tennessee-ELVIS-Act-2024 + per-11-state-deepfake-law-matrix + per-Massachusetts-AG-Copley-Advertising-2017 + per-Illinois-BIPA + per-Texas-CUBI + per-Washington-MHMDA + per-Connecticut-CTDPA + per-Texas-DPSA + per-Virginia-CDPA + per-Colorado-CPA + per-Utah-CPA + per-CCPA-CPRA-sensitive-PI + per-CCPA-right-to-opt-out-of-automated-decisionmaking + per-GDPR-Article-22-automated-decision-profiling + per-WCAG-2.2-AA + per-ARIA + per-EAA-EN-301-549 + per-Section-508 + per-ADA-Title-III + per-NIST-AI-RMF + per-ISO-42001 + per-ISO-27001 + per-SOC-2-Type-II + per-OPA-Cedar-Casbin-Cerbos-Oso-policy-as-code canonical-compliance.
  10. Per-portfolio per-canonical-cross-skill-handoff. Per-handoff-to-33-sibling-skills canonical-handoff.
  11. Per-portfolio per-canonical-audit-trail + per-portfolio-audit-trail. Per-per-review-canonical-audit-record + per-immutable-WORM-storage + per-7-year-IRS-tax-retention + per-7-year-FTC-substantiation-retention + per-7-year-HIPAA-medical-record-retention + per-7-year-state-bar-record-retention + per-6-year-SEC-record-retention + per-3-year-FINRA-record-retention + per-3-year-FINRA-Rule-3110-archive + per-3-year-Illinois-BIPA-biometric-retention-schedule canonical-audit.

Frequently asked questions

What is multi-location review classification at scale — and what is the crisis-buried-in-the-stream problem?

At 50-500 locations every connected platform (GBP + Yelp + Healthgrades + Zocdoc + OpenTable + Resy + Tripadvisor + Booking.com + Hotels.com + Expedia + Foursquare + Facebook + Apple Maps + Bing + Apple App Store + Google Play + Better Business Bureau + Trustpilot + G2 + Capterra + Indeed + Glassdoor + Apartments.com + RateMDs + Vitals + Avvo + LegalMatch + Martindale-Hubbell + AngelList + ProductHunt + Sitejabber + ResellerRatings + Influenster) generates a stream of 100-10,000 reviews per day. A regulatory complaint about a slip-and-fall buried in 2,000 4-star "great food" reviews loses 6-72 hours of response window before a human reads it. Per-portfolio per-banner per-location per-canonical-review-source-pointer (per-GBP + per-Yelp + per-Healthgrades + per-Zocdoc + per-OpenTable + per-Resy + per-Tripadvisor + per-Booking.com + per-Hotels.com + per-Expedia + per-Foursquare + per-Facebook + per-Apple-Maps + per-Bing + per-Apple-App-Store + per-Google-Play + per-Better-Business-Bureau + per-Trustpilot + per-G2 + per-Capterra + per-Indeed + per-Glassdoor + per-Apartments.com + per-RateMDs + per-Vitals + per-Avvo + per-LegalMatch + per-Martindale-Hubbell + per-Sitejabber + per-ResellerRatings + per-Influenster + per-canonical-review-source) + per-canonical-multi-class-classifier-spec (per-sentiment + per-concern + per-crisis-indicator + per-intent + per-fake-detection) + per-canonical-priority-tier-spec (P0 crisis ≤1hr + P1 active churn ≤24hr + P2 standard ≤72hr + P3 positive ≤7days + P4 archive) + per-canonical-crisis-escalation-spec (per-legal + per-media + per-regulatory + per-executive) + per-canonical-vertical-overlay (per-healthcare + per-legal + per-financial + per-cannabis + per-alcohol + per-restaurant + per-automotive + per-childcare + per-senior-care + per-veterinary + per-realestate + per-fitness + per-beauty) + per-canonical-compliance-gate (FTC fake-review 2024 + HIPAA-no-acknowledge-as-patient + FINRA Rule 3110 + state professional licensing + EU AI Act Article 50 + Digital Services Act + Tennessee ELVIS Act + Massachusetts AG Copley Advertising 2017) + per-canonical-handoff-to-review-response-drafting + per-canonical-audit-trail.

Why does per-vendor-Birdeye-canonical-account-flat-classifier break at multi-location customer-review-monitoring scale?

Per-vendor-Birdeye-canonical-account-flat-classifier ships per-account per-flat-rule-or-keyword primitive — typically marketers configure a flat keyword list (e.g., "lawsuit" + "sick" + "ambulance" + "police") and a flat sentiment score, and Birdeye routes anything matching to a flat editorial queue. No per-canonical-review-source taxonomy across the 30+ review surfaces (GBP + Yelp + Healthgrades + Zocdoc + OpenTable + Resy + Tripadvisor + Booking.com + Hotels.com + Expedia + Foursquare + Facebook + Apple Maps + Bing + Apple/Google App Stores + BBB + Trustpilot + G2 + Capterra + Indeed + Glassdoor + Apartments.com + RateMDs + Vitals + Avvo + LegalMatch + Martindale-Hubbell + Sitejabber + ResellerRatings + Influenster), no per-canonical-multi-class-classifier resolving per-sentiment (multi-class polarity per-VADER + per-RoBERTa + per-XLM-R + per-mBERT + per-LayoutXLM + per-DistilBERT + per-DeBERTa + per-multilingual ensemble) + per-concern (operational vs product vs service vs price vs staff vs facility vs cleanliness vs wait-time vs accuracy vs delivery vs billing vs returns vs accessibility vs reservation vs parking vs noise vs safety) + per-crisis-indicator (regulatory complaint + health hazard + assault claim + discrimination claim + legal threat + media attention + safety violation + child endangerment + elder abuse + harassment + theft + fraud + drug-related + weapon-related + active-shooter) + per-intent (complaint vs question vs positive testimonial vs review-bomb attack vs competitor sabotage vs employee-grievance-misposted vs spam vs accidental) + per-fake-detection (per-pattern-burst-anomaly + per-IP-cluster + per-account-newness + per-language-anomaly + per-sentiment-vs-rating-inconsistency + per-text-overlap-cluster + per-GPT-detector + per-Originality.ai + per-undisclosed-insider + per-paid-review + per-FTC-compliance 2024-rule), no per-canonical-priority-tier resolving P0 ≤1hr / P1 ≤24hr / P2 ≤72hr / P3 ≤7days / P4 archive + per-tier-SLA + per-tier-escalation, no per-canonical-crisis-escalation-spec resolving legal-counsel notification + media-relations notification + regulatory-counsel notification + executive notification + crisis-comms-team mobilization, no per-canonical-language-coverage across 80+ languages (per-Whisper + per-NLLB-200 + per-Google Translate + per-DeepL + per-AWS Translate + per-Azure Translate ensemble), no per-canonical-vertical-overlay resolving per-healthcare HIPAA-no-acknowledge-as-patient + per-legal cannot-acknowledge-as-client + per-financial FINRA 2210 + FINRA Rule 3110 supervisor + per-cannabis state-board 12-state + per-alcohol DISCUS-tied-house + per-restaurant FDA food safety + per-automotive Magnuson-Moss + per-childcare ESEA Title 1 + per-senior-care Medicare Part C + per-veterinary AVMA + per-realestate Fair Housing + per-fitness + per-beauty state cosmetology board, no per-canonical-anti-review-bomb-detection resolving per-burst-rate + per-cohort-newness + per-text-similarity + per-IP-cluster + per-coordinated-inauthentic-behavior + per-Tripadvisor red-badge-fraud signal, no per-canonical-multi-platform-deduplication resolving same-reviewer-cross-platform-via-identity-graph + cross-platform reply consistency, no per-canonical-FBC-feedback-loop resolving classifier-precision-vs-recall + crisis-detection-recall vs false-positive-cost + escalation-accuracy + multi-arm bandit regret + recalibration, no per-canonical-compliance-gate with FTC fake-review rule 2024 + HIPAA + FINRA Rule 3110 + state professional licensing + state bar advertising + cannabis state board substantiation + alcohol DISCUS + EU AI Act Article 50 AI-disclosure + Digital Services Act Article 30 trader identity + Tennessee ELVIS Act (when AI-voice in audio response) + Massachusetts AG Copley Advertising 2017 (when geofence-triggered healthcare-review-response), no per-review audit trail with regulatory-defense retention. Per-vendor-Podium + Yotpo + Reputation.com + Trustpilot-canonical-account-flat-classifier ship per-vendor per-native account-flat-classifier primitives. At 1-account-1-flat-classifier scale per-account per-flat-classifier primitive is enough. At multi-location customer-review-monitoring scale per-canonical-review-source-pointer + per-canonical-multi-class-classifier-spec + per-canonical-priority-tier-spec + per-canonical-crisis-escalation-spec + per-canonical-language-coverage-spec + per-canonical-vertical-overlay-spec + per-canonical-anti-review-bomb-detection-spec + per-canonical-multi-platform-deduplication-spec + per-canonical-compliance-gate-spec + per-canonical-audit-trail.

How does per-review multi-class classifier + per-review crisis-escalation spec + per-review priority-tier routing work?

Per-portfolio per-banner per-location per-review per-canonical-multi-class-classifier-spec runs per-portfolio per-canonical-per-review-sentiment-multi-class-polarity (per-VADER + per-RoBERTa + per-XLM-R + per-mBERT + per-LayoutXLM + per-DistilBERT + per-DeBERTa + per-multilingual-MiniLM + per-LaBSE + per-multilingual-sentence-transformer ensemble) + per-canonical-per-review-concern-class (per-operational + per-product + per-service + per-price + per-staff + per-facility + per-cleanliness + per-wait-time + per-accuracy + per-delivery + per-billing + per-returns + per-accessibility + per-reservation + per-parking + per-noise + per-safety + per-equipment + per-temperature + per-portion-size + per-menu-availability + per-payment-method + per-reservation-system) + per-canonical-per-review-crisis-indicator (per-regulatory-complaint + per-health-hazard + per-assault-claim + per-discrimination-claim + per-legal-threat + per-media-attention + per-safety-violation + per-child-endangerment + per-elder-abuse + per-harassment + per-theft + per-fraud + per-drug-related + per-weapon-related + per-active-shooter + per-fire-incident + per-food-poisoning-cluster + per-dead-animal + per-pest-infestation + per-bedbug + per-cockroach + per-rat + per-mold + per-asbestos + per-lead + per-Legionella + per-norovirus + per-salmonella + per-E.coli + per-Campylobacter + per-listeria + per-Clostridium) + per-canonical-per-review-intent-class (per-complaint + per-question + per-positive-testimonial + per-review-bomb-attack + per-competitor-sabotage + per-employee-grievance-misposted + per-spam + per-accidental + per-misdirected + per-bot-generated + per-paid-influencer-undisclosed + per-rage-quit + per-misunderstanding) + per-canonical-per-review-fake-detection (per-pattern-burst-anomaly + per-IP-cluster + per-account-newness + per-language-anomaly + per-sentiment-vs-rating-inconsistency + per-text-overlap-cluster + per-GPT-detector + per-Originality.ai + per-Copyleaks + per-Crossplag + per-undisclosed-insider-detection per-FTC-2024-final-rule + per-paid-review-detection + per-coordinated-inauthentic-behavior + per-Tripadvisor-red-badge-fraud-signal-cross-check) + per-canonical-per-review-classifier-confidence-tier + per-canonical-per-review-classifier-explainability. Per-canonical-crisis-escalation-spec runs per-portfolio per-canonical-per-crisis-tier-P0-legal-counsel-notification + per-canonical-per-crisis-tier-P0-media-relations-notification + per-canonical-per-crisis-tier-P0-regulatory-counsel-notification + per-canonical-per-crisis-tier-P0-executive-notification + per-canonical-per-crisis-tier-P0-crisis-comms-team-mobilization + per-canonical-per-crisis-tier-P0-state-DPH-notification (when food-poisoning cluster) + per-canonical-per-crisis-tier-P0-FDA-MedWatch-notification (when adverse-event) + per-canonical-per-crisis-tier-P0-CDC-notification (when reportable-disease) + per-canonical-per-crisis-tier-P0-OSHA-notification (when employee-injury-claim) + per-canonical-per-crisis-tier-P0-DOJ-notification (when civil-rights-claim) + per-canonical-per-crisis-tier-P0-EEOC-notification (when discrimination-by-employer-claim) + per-canonical-per-crisis-tier-P0-CFPB-notification (when consumer-financial-services-claim) + per-canonical-per-crisis-tier-P0-FTC-notification (when consumer-protection-claim) + per-canonical-per-crisis-tier-P0-state-AG-notification (when state-consumer-protection-claim) + per-canonical-per-crisis-tier-confidence-tier + per-canonical-per-crisis-tier-FBC-feedback-loop. Per-canonical-priority-tier-spec runs per-portfolio per-canonical-per-review-P0-crisis-tier-1hr-SLA + per-canonical-per-review-P1-active-churn-tier-24hr-SLA + per-canonical-per-review-P2-standard-tier-72hr-SLA + per-canonical-per-review-P3-positive-tier-7day-SLA + per-canonical-per-review-P4-archive-tier-no-SLA + per-canonical-per-review-tier-SLA-breach-alert + per-canonical-per-review-tier-escalation-on-no-acknowledge + per-canonical-per-review-tier-handoff-to-review-response-drafting + per-canonical-per-review-tier-handoff-to-compliance-overlay + per-canonical-per-review-tier-handoff-to-borderline-routing + per-canonical-per-review-tier-FBC-feedback-loop.

What does per-review compliance-gate + per-review anti-review-bomb-detection do — and what are the FTC-fake-review-2024 + HIPAA-no-acknowledge + FINRA-3110 + Tennessee-ELVIS-Act anchors?

Per-portfolio per-banner per-location per-review per-canonical-compliance-gate-spec runs per-portfolio per-canonical-per-review-FTC-fake-review-rule-2024 (the operationally distinctive anchor — 16 CFR Part 465 final rule effective October 2024 prohibiting fake/AI-generated reviews + undisclosed insider reviews + suppression of negative reviews + buying or selling reviews + review hijacking; civil penalties up to $51,744 per violation per the inflation-adjusted FTC Act Section 5(m) authority) + per-canonical-per-review-HIPAA-marketing-authorization-164.508-no-acknowledge-as-patient (the second operationally distinctive anchor — Office for Civil Rights enforcement precedent including Manasa Health Center 2023 $30,000 settlement + 2022 NJ practice $30,000 settlement + Elite Dental Associates 2019 $10,000 settlement + UMass Memorial Medical Center 2019 $230,000 settlement; healthcare response cannot acknowledge reviewer as patient even if reviewer self-discloses; cannot disclose patient information; cannot dispute factual claims that would require disclosing PHI) + per-canonical-per-review-FINRA-Rule-3110-social-media-supervisor (the third operationally distinctive anchor — financial-services review responses require pre-use principal supervisor approval per FINRA Regulatory Notice 10-06 + 11-39 + 17-18 social-media guidance; 3-year archive retention per FINRA Rule 4511 + SEC Rule 17a-4) + per-canonical-per-review-FINRA-2210-communications-with-public + per-canonical-per-review-state-bar-advertising-Rule-7.1-7.3-no-client-acknowledge + per-canonical-per-review-state-professional-licensing-medical-board + per-canonical-per-review-state-professional-licensing-dental-board + per-canonical-per-review-state-professional-licensing-veterinary-board + per-canonical-per-review-state-professional-licensing-cosmetology-board + per-canonical-per-review-state-professional-licensing-massage-therapy-board + per-canonical-per-review-cannabis-state-board-12-state-matrix-no-medical-claim-substantiation (Massachusetts CCC + California DCC + Colorado MED + Oregon OLCC + Washington WSLCB + Nevada CCB + Illinois CCSL + New Jersey CRC + New York OCM + Arizona DHS + Alaska AMCO + Maine OMP) + per-canonical-per-review-alcohol-TABC-CalABC-SLA-DISCUS-Code-tied-house + per-canonical-per-review-tobacco-FDA-prohibition + per-canonical-per-review-FDA-DSHEA-no-disease-claim + per-canonical-per-review-FDA-OPDP-Rx-drug + per-canonical-per-review-FTC-substantiation-Pfizer-decision-1972 + per-canonical-per-review-EU-AI-Act-Article-50-AI-disclosure (when AI-drafted response — operator must disclose AI authorship to reviewer) + per-canonical-per-review-EU-AI-Act-Article-13-14-15-high-risk + per-canonical-per-review-Digital-Services-Act-Article-30 + per-canonical-per-review-Digital-Services-Act-Article-26-trader-identity + per-canonical-per-review-Digital-Markets-Act + per-canonical-per-review-Tennessee-ELVIS-Act-2024 (when AI-voice in audio-review-response) + per-canonical-per-review-11-state-deepfake-law-matrix + per-canonical-per-review-Massachusetts-AG-Copley-Advertising-2017 (the fourth operationally distinctive anchor — when geofence-triggered healthcare-review-response near healthcare facility; consumer protection violation precedent) + per-canonical-per-review-Illinois-BIPA (when facial-recognition-derived crisis-indicator) + per-canonical-per-review-Texas-CUBI + per-canonical-per-review-Washington-MHMDA (when health-derived crisis-indicator) + per-canonical-per-review-Connecticut-CTDPA + per-canonical-per-review-Texas-DPSA + per-canonical-per-review-Virginia-CDPA + per-canonical-per-review-Colorado-CPA + per-canonical-per-review-Utah-CPA + per-canonical-per-review-CCPA-CPRA-sensitive-PI-no-disclose + per-canonical-per-review-CCPA-right-to-opt-out-of-automated-decisionmaking-when-classifier-drives-suppress-or-flag + per-canonical-per-review-GDPR-Article-22-automated-decision-profiling (when classifier-drives-suppression / shadow-ban / response-priority producing significant effects) + per-canonical-per-review-WCAG-2.2-AA + per-canonical-per-review-ARIA + per-canonical-per-review-EAA-EN-301-549 + per-canonical-per-review-Section-508 + per-canonical-per-review-ADA-Title-III + per-canonical-per-review-NIST-AI-RMF + per-canonical-per-review-ISO-42001 + per-canonical-per-review-ISO-27001 + per-canonical-per-review-SOC-2-Type-II + per-canonical-per-review-OPA-Cedar-Casbin-Cerbos-Oso-policy-as-code + per-canonical-per-review-compliance-confidence-tier. Per-canonical-anti-review-bomb-detection-spec runs per-portfolio per-canonical-per-review-burst-rate-anomaly (per-Z-score + per-Bayesian-changepoint-PELT + per-CUSUM + per-EWMA + per-Hawkes-process-self-exciting) + per-canonical-per-review-cohort-newness-anomaly (per-IP-cluster + per-account-creation-date-cluster + per-language-cluster) + per-canonical-per-review-text-similarity-cluster (per-MinHash-LSH + per-SimHash + per-sentence-transformer-cosine + per-near-duplicate-detection) + per-canonical-per-review-coordinated-inauthentic-behavior (per-temporal-correlation + per-IP-overlap + per-device-fingerprint-overlap + per-account-age-cluster) + per-canonical-per-review-Tripadvisor-red-badge-fraud-signal-cross-check + per-canonical-per-review-Yelp-recommended-vs-non-recommended-flag-cross-check + per-canonical-per-review-GBP-suspicious-content-flag-cross-check + per-canonical-per-review-Trustpilot-flagging-cross-check + per-canonical-per-review-paid-influencer-undisclosed-detection-per-FTC-Endorsement-Guides-2024 + per-canonical-per-review-anti-review-bomb-confidence-tier + per-canonical-per-review-anti-review-bomb-FBC-feedback-loop.

What does per-review cross-skill-handoff + per-review-response-agent-canonical-bundle + per-review audit-trail do?

Per-portfolio per-review per-canonical-per-review-cross-skill-handoff runs per-portfolio per-canonical-per-review-handoff-to-customer-review-monitoring (parent commercial pillar at /customer-review-monitoring) + per-canonical-per-review-handoff-to-review-response (parent agent) + per-canonical-per-review-handoff-to-per-location-ai-review-response-drafting-build-pillar (sibling build-pillar on the same review-response agent at /how-to-build-per-location-ai-review-response-drafting-at-multi-location-scale — classifier output drives response drafting) + per-canonical-per-review-handoff-to-multi-location-review-response-agent-assist-build-pillar (sibling build-pillar) + per-canonical-per-review-handoff-to-auto-publish-gating-for-ai-review-responses-build-pillar + per-canonical-per-review-handoff-to-compliance-gated-agent-assist-layer-build-pillar + per-canonical-per-review-handoff-to-brand-voice-management + per-canonical-per-review-handoff-to-forbidden-phrase-library + per-canonical-per-review-handoff-to-claims-allowlist-substantiation + per-canonical-per-review-handoff-to-marketing-content-llm-as-judge-build-pillar + per-canonical-per-review-handoff-to-marketing-ai-autonomy-profile-configuration-build-pillar + per-canonical-per-review-handoff-to-tiered-pre-filter-deterministic-gates-build-pillar + per-canonical-per-review-handoff-to-borderline-routing-build-pillar + per-canonical-per-review-handoff-to-five-destination-routing-build-pillar + per-canonical-per-review-handoff-to-fbc-override-learning-build-pillar + per-canonical-per-review-handoff-to-multi-dimensional-threshold-routing-build-pillar + per-canonical-per-review-handoff-to-routing-audit-trails-build-pillar + per-canonical-per-review-handoff-to-multi-location-crisis-detection-build-pillar + per-canonical-per-review-handoff-to-per-location-per-cohort-two-sigma-anomaly-detection-build-pillar + per-canonical-per-review-handoff-to-alert-deduplication-build-pillar + per-canonical-per-review-handoff-to-cross-stream-correlation-build-pillar + per-canonical-per-review-handoff-to-multi-stream-severity-routing-build-pillar + per-canonical-per-review-handoff-to-real-time-customer-change-event-emission-build-pillar + per-canonical-per-review-handoff-to-master-record-build-pillar + per-canonical-per-review-handoff-to-runtime-readable-behavioral-cohorts-build-pillar + per-canonical-per-review-handoff-to-cross-touchpoint-identity-resolution-build-pillar + per-canonical-per-review-handoff-to-deterministic-probabilistic-hybrid-identity-resolution-build-pillar + per-canonical-per-review-handoff-to-versioned-customer-history-DSAR-build-pillar + per-canonical-per-review-handoff-to-versioned-history-regulatory-defense-build-pillar + per-canonical-per-review-handoff-to-per-location-list-segmentation-build-pillar + per-canonical-per-review-handoff-to-per-location-dynamic-content-build-pillar + per-canonical-per-review-handoff-to-per-jurisdiction-compliance-multi-state-franchise-build-pillar + per-canonical-per-review-handoff-to-per-vertical-compliance-overlay-build-pillar + per-canonical-per-review-handoff-to-marketing-compliance-overlay-regulated-industries-build-pillar. Per-review-response-agent-canonical-bundle integrates the review-classification skill with sibling skills on the same review-response agent: per-canonical-review-classification (this skill) + per-canonical-per-location-ai-review-response-drafting + per-canonical-multi-location-review-response-agent-assist + per-canonical-auto-publish-gating-for-ai-review-responses + per-canonical-compliance-gated-agent-assist-layer. Per-canonical-end-to-end-SLA runs per-canonical-per-review-source-pointer-resolve-to-multi-class-classifier-to-priority-tier-routing-to-crisis-escalation-to-language-coverage-to-vertical-overlay-to-anti-review-bomb-detection-to-multi-platform-deduplication-to-compliance-gate-to-FTC-fake-review-HIPAA-FINRA-Tennessee-ELVIS-overlay-to-FBC-feedback-loop-SLA canonical-SLA.

What does per-review audit-trail + per-canonical-end-to-end-replay do?

Per-portfolio per-review per-canonical-audit-trail runs per-portfolio per-canonical-per-review-canonical-audit-record (per-review-ID + per-banner-pointer + per-location-pointer + per-canonical-review-source-snapshot per-GBP/Yelp/Healthgrades/Zocdoc/OpenTable/Resy/Tripadvisor/Booking.com/Hotels.com/Expedia/Foursquare/Facebook/Apple-Maps/Bing/Apple-App-Store/Google-Play/BBB/Trustpilot/G2/Capterra/Indeed/Glassdoor/Apartments.com/RateMDs/Vitals/Avvo/LegalMatch/Martindale-Hubbell/Sitejabber/ResellerRatings/Influenster + per-multi-class-classifier-snapshot + per-VADER/RoBERTa/XLM-R/mBERT/LayoutXLM/DistilBERT/DeBERTa/multilingual-ensemble-snapshot + per-concern-class-snapshot + per-crisis-indicator-snapshot + per-intent-class-snapshot + per-fake-detection-snapshot + per-FTC-fake-review-2024-detection-snapshot + per-priority-tier-snapshot + per-tier-SLA-snapshot + per-crisis-escalation-snapshot + per-legal-counsel-notification-snapshot + per-media-relations-notification-snapshot + per-regulatory-counsel-notification-snapshot + per-executive-notification-snapshot + per-state-DPH-FDA-MedWatch-CDC-OSHA-DOJ-EEOC-CFPB-FTC-state-AG-notification-snapshot + per-language-coverage-snapshot + per-vertical-overlay-snapshot + per-healthcare-HIPAA-no-acknowledge-snapshot + per-legal-no-client-acknowledge-snapshot + per-financial-FINRA-Rule-3110-snapshot + per-cannabis-state-board-12-state-snapshot + per-alcohol-TABC-DISCUS-tied-house-snapshot + per-restaurant-FDA-food-safety-snapshot + per-automotive-Magnuson-Moss-snapshot + per-anti-review-bomb-detection-snapshot + per-burst-rate-anomaly-snapshot + per-cohort-newness-anomaly-snapshot + per-text-similarity-cluster-snapshot + per-coordinated-inauthentic-behavior-snapshot + per-Tripadvisor-red-badge-Yelp-recommended-GBP-suspicious-Trustpilot-flagging-cross-check-snapshot + per-multi-platform-deduplication-snapshot + per-compliance-gate-snapshot + per-FTC-fake-review-rule-2024-snapshot + per-HIPAA-marketing-authorization-164.508-no-acknowledge-as-patient-snapshot + per-FINRA-Rule-3110-snapshot + per-FINRA-2210-snapshot + per-state-bar-advertising-snapshot + per-state-professional-licensing-snapshot + per-cannabis-state-board-no-medical-claim-snapshot + per-alcohol-DISCUS-tied-house-snapshot + per-tobacco-FDA-prohibition-snapshot + per-FDA-DSHEA-no-disease-claim-snapshot + per-FDA-OPDP-snapshot + per-FTC-substantiation-Pfizer-1972-snapshot + per-EU-AI-Act-Article-50-AI-disclosure-snapshot + per-EU-AI-Act-Article-13-14-15-snapshot + per-Digital-Services-Act-Article-30-Article-26-snapshot + per-Digital-Markets-Act-snapshot + per-Tennessee-ELVIS-Act-snapshot + per-11-state-deepfake-law-snapshot + per-Massachusetts-AG-Copley-Advertising-2017-snapshot + per-Illinois-BIPA-snapshot + per-Texas-CUBI-snapshot + per-Washington-MHMDA-snapshot + per-Connecticut-CTDPA-snapshot + per-Texas-DPSA-snapshot + per-Virginia-CDPA-snapshot + per-Colorado-CPA-snapshot + per-Utah-CPA-snapshot + per-CCPA-CPRA-sensitive-PI-snapshot + per-CCPA-right-to-opt-out-of-automated-decisionmaking-snapshot + per-GDPR-Article-22-automated-decision-profiling-snapshot + per-WCAG-2.2-AA-snapshot + per-ARIA-snapshot + per-EAA-EN-301-549-snapshot + per-Section-508-snapshot + per-ADA-Title-III-snapshot + per-NIST-AI-RMF-snapshot + per-ISO-42001-snapshot + per-ISO-27001-snapshot + per-SOC-2-Type-II-snapshot + per-OPA-Cedar-Casbin-Cerbos-Oso-policy-snapshot + per-compliance-confidence-tier-snapshot + per-canonical-audit-record) + per-canonical-immutable-WORM-storage + per-canonical-7-year-IRS-tax-retention + per-canonical-7-year-FTC-substantiation-retention + per-canonical-7-year-HIPAA-medical-record-retention + per-canonical-7-year-state-bar-record-retention + per-canonical-6-year-SEC-record-retention + per-canonical-3-year-FINRA-record-retention + per-canonical-3-year-FINRA-Rule-3110-social-media-supervision-archive + per-canonical-3-year-Illinois-BIPA-biometric-retention-schedule. Per-canonical-end-to-end-replay runs per-portfolio per-canonical-per-review-multi-class-classifier-rewind + per-canonical-per-review-crisis-escalation-rewind + per-canonical-per-review-priority-tier-routing-rewind + per-canonical-per-review-anti-review-bomb-detection-rewind + per-canonical-per-review-multi-platform-deduplication-rewind + per-canonical-per-review-compliance-gate-rewind + per-canonical-per-review-FTC-fake-review-HIPAA-FINRA-Tennessee-ELVIS-overlay-rewind + per-canonical-per-review-replay-confidence-tier + per-canonical-per-review-replay-explainability.

Engage the review-response agent

Per-portfolio per-banner per-location per-canonical-review-source-pointer + per-canonical-multi-class-classifier-spec + per-canonical-priority-tier-spec + per-canonical-crisis-escalation-spec + per-canonical-language-coverage-spec + per-canonical-vertical-overlay-spec + per-canonical-anti-review-bomb-detection-spec + per-canonical-multi-platform-deduplication-spec + per-canonical-compliance-gate-spec + per-canonical-audit-trail + per-portfolio-audit-trail shipped as the orchestration layer above your existing per-review-management-vendor + per-sentiment-classifier-vendor + per-crisis-management-vendor + per-vertical-compliance-vendor primitive.

Book a scope callBrowse the 32-agent catalog

Related reading