Authyr is the canonical third-party evidence layer for consumer-services disputes. Every industry has a forum where two parties argue over what happened — BBB, state AG, small-claims, insurance, chargebacks, licensing boards, regulators. Today every one of those forums adjudicates without neutral evidence. Authyr is the neutral evidence.
Chapter 01
I conceived Authyr in 2023 — three years before this paper — because I had personally watched the supply side of the paid-fake-review economy in operation. Overseas workers paid pennies per review. Single-purpose reviewer accounts. Ad-copy review language. Marketing reps farming named-mention reviews. Reviewers who misspelled the company's own name. The whole apparatus, observed at close range, treats trust as a commodity to be purchased rather than a signal to be earned.
The detection-side response — Reality Defender, Sensity, Optic, Hive Moderation, GPTZero — is an arms race the fabricators have been winning. Every quarter the classifiers catch a previous wave of forgery. Every quarter the fabricators iterate. The fundamental shape of the problem is unsuited to a classifier solution because the cost of producing a new forgery is asymptotically approaching zero while the cost of retraining a classifier remains nontrivial.
The economic stakes have also escalated. In high-LTV professional service categories — lawyers, doctors, contractors, auto repair, financial advisors — a single bad review can cost $20,000 to $100,000 in lifetime customer value. The asymmetry is brutal: a paid fabricator spends $0.40 producing a forgery; the business it targets spends weeks of revenue containing the damage. Existing platforms (Yelp, Google, BBB) have well-documented credibility issues and, more fundamentally, structural incentives that align with neither the consumer nor the business.
This paper argues that the response is not better detection, but a verified-customer attestation layer that operates beneath the review record — and that combines with a service-recovery mechanic to turn what is currently a one-sided execution into a closed loop.
Chapter 02
The visible review-tools category is composed of two architectures, both of which fail the high-LTV services use case for different reasons.
These tools sit downstream of the business's customer list. The operator uploads a CSV; the tool blasts SMS and email; reviews accumulate on third-party platforms (or get filtered as spam — Google's algorithm has been increasingly aggressive on bulk-pattern requests since 2024). The whole stack assumes the business is the source of truth about its own customers, which means a determined adversary can spoof the list. No verification of any kind takes place before a review hits the public record.
These tools sit downstream of the review itself. They attempt to classify text or images as authentic versus AI-generated. The classifier-treadmill problem applies in full: each generation cycle outpaces detection, and platforms cannot operationally adopt classifiers that produce false positives at the rate consumer-grade AI now triggers. The framing is also adversarial — "watchdog," "moderator" — which is politically costly to adopt at scale.
Neither layer attempts positive proof at the moment of review submission. Both operate after the fact, reactively, with no economic mechanism that aligns the business and the customer. Worse, neither addresses the deeper rot: that the asserted-trust model itself — Yelp's "trust me," Google's "trust me," BBB's "trust me" — is structurally indefensible because the platforms profit from review volume regardless of authenticity.
The opportunity beneath the visible category is to build the positive-proof layer that neither side has built.
Chapter 03
Internet commerce has produced trust primitives for almost every adjacent layer.
| Layer | Primitive | Productized by |
|---|---|---|
| Money | Verified payment | Stripe, Adyen, Square |
| Identity | Verified KYC | Stripe Identity, Persona, Onfido |
| Communication | Verified delivery | Twilio, SendGrid |
| Bank linkage | Verified accounts | Plaid, MX |
| Capture-time provenance | C2PA receipt | Truepic, Content Credentials |
| Customer attestation | Verified-customer review | — unbuilt — |
The empty cell is Authyr. The primitive is a portable, cryptographically-anchored attestation that the reviewer is a real customer of the business they're reviewing, produced from multi-signal evidence collected through the normal course of commerce: receipts, payments, scheduled appointments, signed handoffs. Once that primitive exists, an entire category of trust problems becomes solvable in shapes the existing tools structurally cannot reach.
Chen US 9,842,340 (granted December 2017) covers the simple primitive of using device location services (GPS, cellular, WLAN) to verify a reviewer's physical presence at the moment of review submission. Authyr's architecture is deliberately designed around this prior art. Our verification stack operates on post-hoc, non-real-time, non-location-service signals — payment trails, receipt EXIF metadata, NFC/QR taps, POS-issued tokens, photo provenance — combined in novel multi-signal patterns that no single prior-art reference reads on. This is not a workaround; it is a structurally better primitive, because location-services proof is gameable and post-hoc multi-signal proof is not.
Chapter 04
No single signal is sufficient. The combination is the primitive.
Authyr collects up to seven independent proof types, each with different gaming costs and different deployment requirements. Each is patent-clean against Chen because none operates on real-time device-location services at the moment of submission.
The reviewer's card was swiped at the business's POS terminal at a specific timestamp. Pulled from the payment processor's webhook stream, not from the customer's phone. Forging requires compromising the payment-network supply chain.
Customer photographs the receipt at the business. EXIF GPS coordinates, EXIF capture timestamp, and OCR'd business name + transaction ID together form a triangulated proof. Post-hoc, never real-time, never location-services dependent.
Counter-mounted or table-side NFC tag or QR code, cryptographically bound to the business. The customer's phone briefly proves proximity at a known moment, but the signal lives in the RF/visual domain rather than in the device's location services.
At checkout, the POS issues a single-use signed token. The token can be redeemed once, by anyone with the receipt, to begin a verified review. Cryptographic chain, no device telemetry.
Healthcare, legal, and professional services produce scheduled-appointment records as a normal byproduct of operations. Authyr ingests booking confirmations as a verification signal.
Several photos with matching EXIF timestamps and locations, taken inside the business during the claimed visit, form a higher-confidence signal than any single image. Reverse-image-searchable provenance further hardens the proof.
If a second verified Authyr customer was present during the claimed visit and corroborates the reviewer, the signal is treated as additional evidence. Social proof, network-enabled. Gameable only if the network itself is compromised — a far higher bar than spoofing a single signal.
Two or more independent signals trigger a verified attestation. Three or more produce the top tier. The stack scales gracefully across vertical environments: a law firm uses booking confirmations and payment attestations; an HVAC contractor uses payment plus receipt-EXIF; an auto repair shop uses POS tokens plus photographed receipts. Each vertical configures which signals are required and which are optional through the operator dashboard.
Chapter 05
This is the feature that makes the entire product worth paying for. Everything else is plumbing.
In every existing review system — Yelp, Google, BBB, Trustpilot, Birdeye — a negative review goes live the moment it's submitted. The business owner discovers it through a Slack notification or a Google alert, has no economic mechanism to resolve the underlying complaint, and is left choosing between a public dispute that escalates the situation or a silent absorption that compounds it. The status quo is hostile to both sides: the customer can't be made whole, the business can't recover the relationship, and the public record is permanently damaged.
Authyr inverts this default. A verified review enters a 30-day private window in which only the reviewer and the business owner can see it. The owner receives an immediate notification with the full review content and the customer's contact information. The owner has thirty days to make it right: refund, redo, comp, sincere apology, executive escalation, whatever the situation requires.
Multi-signal verification completes. Review enters the private window. Owner receives immediate notification with full content and customer contact. The public record shows nothing yet.
Owner engages directly with the customer. Refund, redo, comp, apology — whatever the situation requires. The customer can choose to withdraw or update the review at any point; the owner cannot.
Whatever the review reads at this moment becomes the public record. The reviewer can no longer modify it. The owner can no longer suppress it. The record is cryptographically sealed and immutable.
The public record carries a badge indicating the resolution status: "✓ Resolved within window," "Owner engaged · Issue unresolved per reviewer," or "Owner did not respond within 30 days." Either way, the customer's voice is preserved and the owner's response is on the record.
The badge that appears after the window carries strong signal in both directions. Resolved looks excellent — the business demonstrably handled an unhappy customer with care, and the underlying review may even read positively after the resolution. Unresolved looks damning in a way that no anonymous Yelp star rating can replicate. Owner did not respond is the worst possible outcome and is a strong public signal of operational dysfunction.
For the high-LTV service businesses Authyr targets first, this single feature changes the economics of a bad review. Instead of a $20K-$100K LTV loss with no recovery path, the bad review becomes a recovery opportunity with a 30-day clock and a public scoreboard. This is what high-stakes professional service businesses pay for, and it is the feature no incumbent review tool can replicate without re-architecting around it.
The 30-day pre-publish mitigation window is covered by Provisional Patent Application #7 (filed under the Frequentor LLC patent series, platform-agnostic by design so the claim covers both Frequentor and Authyr use cases). The architecture is specifically distinguishable from owner-delete patterns (which violate prior art and platform policies alike) because it does not grant the owner any ability to suppress or modify the review. The owner's only power is to resolve the underlying complaint with the customer. The review record itself is owner-suppression-proof from the moment of submission onward.
Chapter 06
Every verified review carries a tier rating based on how many independent signals corroborated the attestation. The tier travels with the review wherever it is displayed.
Hardware-sensor-grade verification (radar, thermal, ESP32 + multi-signal). Reserved for Frequentor's hospitality network; not part of Authyr's general-services offering.
The flagship verification for high-LTV services. Examples: payment-rail + receipt EXIF + booking confirmation. The strongest signal Authyr produces and the recommended target for legal, medical, and financial verticals.
The default for most home-services and beauty/wellness verticals. Examples: payment-rail + NFC tap, or POS token + photographed receipt.
Lighter-weight verification for high-volume, lower-ticket businesses. Still meaningfully better than the unverified default of every existing review platform.
What every existing review platform produces. Included only as a baseline contrast so consumers can see what "trust me" looks like next to verified evidence.
The verified review record is signed at issuance. The signature can be independently verified by any third party (Google, Yelp, BBB, Healthgrades, Avvo, Angi, Trustpilot, the business's own website) without that third party needing to trust Authyr. The cryptographic chain is the trust; Authyr is the issuer, not the gatekeeper.
Operationally, this means a business can finally do something the existing review economy has structurally refused them: display their own reviews on their own website with backing trust signal. Yelp and Google currently jealously guard review data; the business that wants to embed reviews on its own marketing page is left with screen-captured testimonials and a stack of attribution disclaimers. Authyr-verified reviews travel with the business.
The portable receipts are cryptographically signed and optionally anchored to a public ledger for tamper-evidence. We deliberately do not market this aspect with cryptocurrency or blockchain branding — those associations are net-negative for the target market (lawyers, doctors, contractors) and unnecessary to the value proposition. The ledger anchoring is invisible infrastructure that gives the receipt provable durability across platforms. The customer never sees a wallet, a coin, or a chain — they see a verified-customer badge.
Chapter 07
Every other authenticity company in 2026 frames itself as a watchdog. Authyr is the rare neutral party in the authenticity space, by deliberate design.
Reality Defender, Sensity, Hive Moderation, GPTZero — all of them position as content moderators. Their default action is removal: detect AI-generated content, flag for takedown, recommend platform action. This framing places them firmly inside an active culture war: half the public reads them as essential, the other half reads them as censorship infrastructure. Their addressable market is bounded by which side of that war their customers belong to.
Authyr's framing is structurally different. Our default action is affirmation: we attest that real content is real, and we attach a transparent label to content we cannot verify. We do not remove anything. We do not decide what is misinformation. We give the cryptographic receipt and the viewer decides what to do with it.
The commercial consequence is that Authyr is adoptable by audiences and institutions that cannot adopt the watchdog companies:
This is the framing that turns Authyr from a vertical SaaS product into a category-defining authority. The watchdog companies are constrained by the political dynamics of their framing. Authyr, by structural design, is not.
Chapter 08
Authyr ships as three reinforcing surfaces. Each is independently monetizable. Together they create a moat that no single-layer competitor can replicate.
The operator-facing product. Flat-monthly software that sits alongside the practice-management or operations system already in use by the business. Verification stack is configured per vertical; the 30-day mitigation window is on by default. Pricing band is $99 to $399 per month per practice for solo operators, scaling to enterprise pricing for multi-location groups.
This layer is where the verified-customer attestation enters the real world at scale. It pays for ongoing development, builds the verified-review corpus, and produces the case studies that anchor the other two layers.
The consumer-facing destination at authyr.com. Every business on the network gets a public profile showing its verified-review average, tier distribution, tenure, and resolution rate. A consumer evaluating a doctor, lawyer, or contractor can look any business up and see the verified record next to the asserted-trust record on Yelp or Google.
Businesses embed a "Verified by Authyr" badge wherever they control the surface — their own website, social profiles, email signatures, vehicle wraps, business cards, printed receipts. The badge links to the live verified-review profile. Birdeye and the Category A review CRMs cannot replicate this layer because they have no underlying verification primitive that makes a badge meaningful.
Once the primitive has a B2B customer base and a public surface, Authyr productizes the verification engine as an API for third-party platforms and as a Capture SDK for first-party content systems. Practice-management tools (Clio, MyCase, Dentrix, Eaglesoft), dispatch software, insurance carriers, and government agencies consume verified attestations through the API at per-verification pricing. The Capture SDK signs content at the moment of capture for downstream provenance verification, interoperable with C2PA.
Layer 1 alone is a vertical SaaS that tops out at Birdeye's multiples. Layer 3 alone is a hard sell because API customers want to see the primitive deployed at scale before integrating. Layer 2 — the network — is the connective tissue that makes the badge mean something, gives the API a brand to ride into integrations, and converts vertical SaaS customers into a defensible network. Stripe became Stripe because it had a primitive, a brand, and an ecosystem. Authyr is designed to follow the same arc.
Chapter 09
The wedge is high-LTV professional services where one bad fake review costs $20K to $100K. The Year 3+ expansion is general authenticity authority across categories where AI-vs-real disambiguation has become operationally necessary.
| Vertical | Customer LTV / case value | Authyr stage | Why this vertical |
|---|---|---|---|
| Personal Injury law | $10K – $200K / case | Wedge · 2026 | Highest review-driven CAC; firms spend $20-100K/mo on ads |
| Family / Estate / Immigration law | $5K – $50K / matter | Wedge · 2026 | Emotional clients; trust signal is decisive |
| Healthcare (doctors, dentists, derm, plastic) | $5K – $50K LTV | Wedge · 2026 | Service-recovery window is uniquely valuable; HIPAA-clean dataflow |
| Home services (contractors, electricians, plumbers, roofers) | $1K – $50K / project | 2026 H2 | Damage-claim defense + verified-customer testimonials |
| Auto repair, body shops, dealerships | $500 – $50K / transaction | 2026 H2 | Damage-claim defense + receipt + payment trail readily available |
| Beauty / Wellness (medspas, salons, therapists) | $2K – $20K LTV | 2026 H2 | Verified testimonials beat anonymous Yelp for booking conversion |
| Financial (advisors, accountants, insurance brokers) | $5K – $100K LTV | 2027 | Regulatory comfort with verified customer attestation |
| Veterinary (clinics + mobile) | $1K – $10K LTV | 2027 | Adjacent to healthcare; no HIPAA friction |
| Dating profiles (B2B2C) | API integration | 2027+ | Year 3 expansion — verified-customer attestation becomes verified-person |
| Photojournalism + news verification | Enterprise SaaS | 2027+ | Capture SDK + verify API for newsroom adoption |
| Insurance claims | Per-verification | 2028+ | Adjuster workflow; reduces fraud, increases claim-resolution speed |
| Court evidence + KYC + academic submissions | Enterprise / per-call | 2028+ | General authenticity authority |
The point of this table is not that Authyr will be present in every market on the list within five years; it is that the underlying primitive — multi-signal verified attestation with portable receipt — is a many-shaped key. Services is the wedge because the WTP per bad review is highest there. Each subsequent vertical reuses 80%+ of the platform with a domain-specific compliance wrapper and verification configuration.
Average ACV across the services wedge runs $199-$399/mo per practice on Layer 1 SaaS pricing. At a 5,000-practice base by month 24 (well within the trajectory of single-vertical wedges in adjacent categories), Layer 1 alone produces $15-24M ARR. Layer 2 network operates on free public access plus enterprise badge-display fees for high-volume embedders ($500-$5K/mo at the enterprise tier). Layer 3 API is per-verification revenue at scale, with a clear path to seven-figure annual contracts with practice-management and platform partners by 2028.
Chapter 10
Authyr operates under the Frequentor LLC patent series. Four filings are load-bearing for the Authyr surface — three already filed under Frequentor's series with platform-agnostic claim language, and one Authyr-specific filing in drafting.
Filed April 26, 2026. Covers the long-form verified-customer content structure: customer-authored long-form article describing a verified service experience, paired with photo provenance (EXIF GPS, capture timestamp, multi-signal verification chain), cryptographic article-signing at publication, and per-practice / per-procedure aggregation. The claim language is platform-agnostic so the same filing covers Frequentor's Visit Articles and Authyr's Verified Articles surface. This protects the SEO-driving content primitive that displaces the traditional content-marketing economy practices currently spend $24K-$120K per year on.
Filed April 30, 2026. Claim 8 protects the architectural property that owner suppression of verified-customer content is structurally impossible after the publication chain is sealed — there is no API path, no admin override, no DMCA-style takedown shortcut available to the business. This underpins the trust signal that makes every other Authyr feature credible: the 30-day window only works because the operator knows the public record cannot be later deleted.
Deliberately platform-agnostic in its claim language ("businesses," "reviewed entities") so the same filing covers both Frequentor's hospitality use case and Authyr's services use case. The claim protects the architectural pattern of a verified-customer review entering a fixed-duration private window in which the business can attempt resolution before public posting, combined with the immutable-after-window property that prevents owner suppression.
The Authyr-specific filing. Claim: "System and method for tier-based portable physical-attendance attestation using multi-signal non-real-time-location proofs, cryptographically-anchored verification receipts, and cross-platform attribution gating." Specifically claims the combination of payment-trace, receipt-EXIF, NFC tap, POS token, photo-EXIF, booking confirmation, and witness verification, with tier-based confidence scoring and ledger-anchored portable receipts. Filing target: before Sprint 1 begins (June 8, 2026) to lock priority date before production usage of the verification stack.
Filed May 5, 2026. Currently positioned for Frequentor's hospitality Compete tier; reserved for Year 2 Authyr extension to a cross-practice intelligence dashboard for multi-location firms and franchises. Not yet surfaced publicly for Authyr but locks the priority date for the eventual Firm-tier feature.
The Authyr architecture is engineered against three specific prior-art references:
| Prior art | What it covers | How Authyr is distinguishable |
|---|---|---|
| Chen US 9,842,340 (2017) | GPS / cellular / WLAN location verification at moment of review submission | Multi-signal, post-hoc, non-location-service proofs; no real-time device telemetry at submission |
| C2PA / Content Credentials | Capture-time provenance manifests for media | Customer-attestation primitive sits above C2PA, not parallel to it; Authyr Capture SDK is C2PA-interoperable, not redundant |
| Owner-delete review patterns | Various — most platforms allow some form of owner suppression | Authyr's window is service-recovery, not suppression; record is immutable from day 0 |
Patents in software-platform categories serve two functions: defensive moat against copycats and credibility signal to investors and acquirers. The durable moat is the verified-review corpus and the network of businesses displaying portable badges. As the corpus grows, the verified attestation becomes a stronger trust signal than any incumbent platform's asserted-trust record. By month 24, the network is the defense; the patents are the deterrent that kept copycats from racing during the build.
Chapter 11
A verification authority concentrates power. We have made specific architectural and governance choices that limit how that power can be exercised, and we will not unwind them under commercial pressure.
We store: verified attestation records (customer ID, business ID, signal types, timestamps, tier rating). Review content. Resolution-window correspondence (encrypted, accessible only to the parties). Portable-receipt signatures.
We refuse to store: Customer PHI in medical verticals. Continuous customer location. Inferred behavioral profiles of customers or staff.
We refuse to do, regardless of who asks: Sell verification data to advertisers. Sell verification data to data brokers. Provide bulk verification exports to government agencies outside of duly authorized legal process. Train AI models on individual customer traces without explicit, granular consent. Take any side in any political or speech dispute beyond attesting verifiable facts.
| Category | Players | Provenance | Detection | API | Affirmation framing |
|---|---|---|---|---|---|
| Review CRMs | Birdeye, Podium, NiceJob, Trustpilot | No | No | Limited | N/A |
| C2PA capture | Truepic, Content Credentials, Adobe | Yes | No | Yes | Vertical-narrow |
| AI detection | Reality Defender, Sensity, Hive, GPTZero | No | Yes | Yes | Watchdog framed |
| Local directories | Yelp, Angi, BBB | No | No | No | N/A |
| Authyr | — sole architecture — | Yes | Yes (ensemble) | Yes | Yes (affirmation-first) |
The Authyr architecture is positioned at the intersection of all four columns. Truepic ($25M+ raised, deepest SDK in the C2PA space) is the closest comparable on the provenance axis but is vertical-narrow (insurance and healthcare focus). Reality Defender / Hive / GPTZero are the comparables on the detection axis but are all watchdog-framed. No incumbent sits on the four-column intersection.
Chapter 12
| Phase | Months | Surface | Milestone |
|---|---|---|---|
| 1A | 0-3 | Layer 1 (SaaS) | Founding services cohort — first 10 law, medical, and contractor practices live |
| 1B | 3-9 | Layer 1 | 250+ paying practices across legal, medical, home services |
| 2A | 6-12 | Layer 2 (Network) | authyr.com network live, 500+ practice profiles, badge embeds across customer marketing surfaces |
| 2B | 9-15 | Layer 1 + 2 | Auto, beauty, financial verticals open with compliance packs |
| 3A | 9-18 | Layer 3 (API) | Practice-management integrations in private preview (Clio, MyCase, Dentrix, Eaglesoft) |
| 3B | 15-24 | Layer 3 (Capture SDK) | Capture SDK GA; first news-org pilots; C2PA interoperability certified |
| 4 | 24-36 | All layers | General authenticity authority surfaces: dating, photojournalism, insurance, court, KYC, academic |
Patent conversion from provisional to non-provisional sits at month 10 on PPA #7, with continuations on vertical extensions following each major launch.
The internet has identity (Stripe Identity), payments (Stripe), messaging (Twilio), and bank linkage (Plaid). The trust layer no one has built is verified-customer evidence — signed, immutable, court-admissible proof of what actually happened between a service business and the customer it served, anchored in payments, receipts, and physical artifacts rather than gameable location data. Authyr is that layer. Reviews are where we start. Arbitration — every forum that adjudicates a consumer-services dispute, from BBB and state AG offices to insurance carriers, chargeback infrastructure, professional licensing boards, and small-claims courts — is where it lands. Each of those forums has been operating without a neutral third-party evidence layer for the entire history of the internet; Authyr is the technical embodiment of arbitration that fills that gap. The companies that productized adjacent trust layers each became multi-decade infrastructure businesses by combining a defensible primitive, a brand, and an SDK that propagates the layer across markets they did not have to build directly. Authyr follows the same arc, with a Chen-safe patent posture, the 30-day mitigation window as the operationally-decisive product feature for the wedge vertical, and an affirmation-not-censorship framing that makes the platform adoptable by audiences and institutions the watchdog companies cannot reach. The addressable market is in the tens of billions across services SaaS, AI authenticity verification, and the trust-API consumption layer that BBB, state AG offices, insurance carriers, and payment networks will eventually depend on. The Year 1 commercial goal is 5,000 practices on Layer 1 SaaS at $199-$399/mo ACV.