CareFirst — The Vision Compendium

The broad CareFirst strategy, working hypotheses, and open questions in one private compendium.

Choose one page language / 选择页面语言

This compendium preserves CareFirst's broad strategy and working hypotheses. It does not define current implementation scope. docs/PRODUCT.md and accepted entries in docs/DECISIONS.md govern current product work.

1. The Central Idea

This document is the single home for the CareFirst idea. It gathers everything Jim has envisioned for CareFirst — across many conversations, in one place — so that nothing gets lost and the full picture can be seen at once. It is written to seed a future agent that will act as the ongoing authority on this idea.

A note on how to read it: CareFirst is early. Most of what follows is vision and design thinking, not settled commitments. Where something is genuinely decided or genuinely urgent, this document says so plainly. Everything else should be treated as the raw material of the idea — preserved in full breadth, open to revision.

CareFirst aims to become patient-controlled infrastructure for lifelong dental identity, records, permissions, care coordination, and trusted outcomes.

It begins with dental implants, for a simple reason: an implant may stay in a patient's mouth for decades, while the information needed to maintain it — which system, which components, which surgeon, which lab — typically stays trapped inside individual dental offices, laboratories, imaging centers, manufacturers, or obsolete software. When that information is lost, the implant becomes a mystery object in a patient's body.

Two sentences anchor everything:

The clinical promise: No implant should ever become unidentified.

The business thesis: The patient — not any one dental office — should become the persistent connection among records, providers, countries, implant components, maintenance, and future care.

2. How the Whole Thing Works — The Trust-First Sequence

CareFirst is one comprehensive ecosystem that serves providers and patients in different ways. The way in is deliberate:

Phase 1 — Earn trust (the "Implant Wallet" era). CareFirst helps dental providers with marketing — getting their practices visible in Google, Google Maps, and AI search engines like ChatGPT and Perplexity, in both English and Chinese. This is valuable in its own right and generates revenue. But its deeper purpose is that every marketing client becomes a doorway: their patients get onboarded into the CareFirst Wallet, and the practice itself is inducted into the ecosystem.

Phase 2 — Grow the ecosystem. Once patients and providers are inside and trust is established, the Wallet grows new capabilities: cross-border care coordination, the provider network, rewards ($FIRST), outcome tracking, and eventually the full platform.

The marketing business is therefore not a side hustle and not the end goal. It is the patient-acquisition and trust-building engine for wallet adoption. Revenue from it is real; the strategic payload is induction into the ecosystem.

3. The Components

The strategic components of the vision are described in the independently switchable sections below.

3.1 The venture and the platform

The company has gone by several names in different conversations — CareFirst, CareFirst AI, CareFirst Tech, and (historically) CareFirst32. No single name is canonical yet, and the trademark findings in Section 8 need to be part of choosing one.

CareFirst OS is the name for the long-term platform architecture — the common infrastructure that all future dental services could share: identity, permissions, records, provenance, intake, workflow, human approval, care coordination, outcomes, payments, and interoperability. Today it is a strategic umbrella — the shape of the ambition — rather than a finished product.

Corporate structure remains open: whether Vitang Technology LLC or a new entity owns and operates CareFirst, founder equity, Samuel's formal role, and the assignment of patents, software, and trademarks all still need to be settled.

3.2 Care Wallet — the central patient product

Every patient gets one lifelong wallet that travels with them across providers, cities, and countries. It holds:

• their identity;
• implant and dental records — X-rays, CBCT, intraoral scans;
• implant identity details (brand, model, size, lot, surgical date, surgeon, abutment compatibility);
• treatment records, written by providers;
• lab and material traceability — which lab made the crown, which alloy, which technician;
• referral history with full clinical handoffs;
• insurance documents;
• authorizations, shares, and a log of who has accessed what;
• multiple CareCases (below);
• future maintenance and outcomes.

The core experience: the patient decides who sees their records, for what purpose, and for how long — with scoped, expiring, revocable sharing. The provider keeps practicing as they always have; the patient carries the continuity.

3.3 CareCase — episodes within a wallet

A wallet contains one identity but many concerns. A CareCase is one episode or concern: preserving records after a move, identifying an existing mystery implant, evaluating missing teeth, a full-arch treatment case, an implant complication, later orthodontic or preventive care.

Stable identity information is reused across cases. Changing facts — medications, symptoms, consent, contact information — are versioned and reconfirmed each time, so nothing stale silently drives a clinical decision.

3.4 Global Implant Passport — the first flagship artifact

The Passport is the first high-value clinical object inside the wallet: a complete, verified record of an implant. It can eventually hold the manufacturer and system, connection and dimensions, lot/serial/UDI numbers, site and placement date, placing and restorative clinicians, abutment/screw/driver information, laboratory and prosthetic records, all imaging, the treatment and maintenance history, complications and repairs, and the record's provenance and verification status.

The key insight: the Passport is not a card. Its value comes from being connected — to the source records, to the patient's permissions, to future providers, and to lifelong maintenance. (See Section 8 for an important naming caution.)

3.5 The Network

The proposed open ecosystem of participants: U.S. reviewing dentists, overseas treatment centers, local aftercare providers, implant specialists, laboratories, imaging centers, manufacturers, distributors, coordinators, insurers, and researchers.

Guiding principle: the Network should not belong to any one clinic, manufacturer, or delivery partner. Its value is neutrality, continuity, and evidence.

Wang Huizhang (王会长), head of a private dental association in China, is an initial collaborator, helping recruit accredited top-tier implant providers (in cities like Qingdao) as early network partners. How the network is ultimately governed — and how any single association's role squares with the neutrality principle — is an open question that hasn't been worked through yet.

3.6 AI Clinical Trust Engine — the future evidence layer

Over time, the network generates longitudinal evidence: documentation completeness, provider credentials, treatment methods, components used, maintenance compliance, complications, patient-reported outcomes, follow-up duration, and treatment-center performance.

The Trust Engine organizes this into a quality layer. One firm design conviction: it should begin as an internal evidence and completeness system. A public "Doctor Trust Score" would be premature and dangerous without risk adjustment, adequate sample sizes, real follow-up periods, confidence intervals, and appeal procedures. Get the evidence right privately before scoring anyone publicly.

3.7 The marketing engine — the front door

The near-term wedge, and the trust-building engine of Phase 1:

• The market: roughly 500 Chinese-American dentists across North America, most with outdated websites, weak local SEO, no presence in AI search, and poor Google Maps rankings — an underserved market where Jim has a warm network advantage.
• The core insight: patients have started asking ChatGPT, Perplexity, and Google's AI "who's the best implant dentist near me" — and these practices are invisible there. That invisibility is the opening.
• The product: GEO (Generative Engine Optimization — getting a practice cited by AI engines, not just ranked on Google), built on three layers: technical infrastructure (AI-crawler permissions, schema markup, Bing Places), answer-first bilingual content targeting real patient questions, and authority building (reviews, directories, Chinese platforms like Xiaohongshu and WeChat).
• The delivery model: a six-agent AI marketing team — Research (AI-visibility monitoring), Website, SEO/GEO content, Conversion (new-patient front desk), Customer Service (recall and follow-up), and Reporting — so one operator can serve many practices. Every agent output is graded: fully automatic for safe tasks, human-approved for publishing, human-written for anything touching medical claims or patients.
• The signature sales move: "let them use their own phone" — the prospect asks ChatGPT themselves and watches competitors appear while they don't.
• Honesty rules, non-negotiable: GEO results take 2–4 months, and pricing and expectations are set accordingly. Never promise that AI will recommend a practice — no one can guarantee that. No fabricated testimonials, ever; no fabricated audit data, ever.

3.8 The 4 Levels of Dentistry — the philosophy

A framework from Jim's book describing the patient journey in five tiers, each with a one-line goal:

Level	Goal
I	Nothing ever breaks
II	Calm the foundation
III	Save what's yours
IV	Rebuild for decades
↺ Return	The loop never opens

Its role in CareFirst: a philosophical underpinning and supporting document — a source of voice, patient education, and content-marketing raw material — rather than a strategic pillar of the business. It should always carry the framing: an educational framework, not a treatment recommendation for any individual.

3.9 Blockchain — the honest position

Blockchain was part of the original vision of patient-controlled dental records. The mature position that has emerged:

• Identifiable clinical data stays securely off-chain. Ordinary identity, encryption, access control, and revocation do the real work of patient control.
• Later, hashes or Merkle roots could be anchored on-chain where independent integrity proof genuinely helps — verifying that a record hasn't been altered, credentialing, settlement, or multi-party governance.
• The clear-eyed truth: blockchain can prove that something hasn't changed. It does not, by itself, create meaningful consent or safe patient control.
• The open question that gates everything: what specific problem actually requires blockchain? Until that question has a concrete answer, blockchain remains part of the vision's vocabulary, not the product's architecture.

3.10 $FIRST — the rewards and incentive concept

Current status, stated plainly: no firm decisions have been made about $FIRST in any direction. It is a future ecosystem concept, not a specified token and not a launched program.

The full breadth of ideation is preserved here, because all of it may prove useful:

• As a token, the design explored a four-stakeholder ecosystem — Patients, Providers, Suppliers, Insurance — each earning and redeeming, with a full supply allocation and roadmap.
• As a loyalty program, the design explored non-transferable airline-miles-style points: earned, never bought; no cash value; redeemable for services; a compliance-friendly on-ramp.
• Possible uses discussed: payment for Wallet and network services; rewards for completing records; maintenance and preventive-care incentives; outcome-reporting incentives; provider compensation; network settlement; staking or credential participation; governance.

Everything remains open: supply, issuance, chain, custody, vesting, governance, transferability, taxation, regulatory classification, and healthcare-referral compliance.

The coherent sequence everyone has converged on: establish real utility first, test incentives with ordinary payments and rewards, and only introduce a transferable token if and when it solves a legitimate problem that ordinary mechanisms can't.

One practical caution: web pages exist for both the token version and the rewards version of $FIRST. They are design explorations, not launched products, and shouldn't be publicly promoted while the concept is undecided.

4. The Cross-Border Care Loop

The larger commercial wedge is a complete loop:

U.S. evaluation → complete records → patient-authorized overseas review → treatment → verified Implant Passport → U.S. aftercare → lifelong maintenance and outcomes.

Concretely: U.S. general practitioners act as local diagnostic nodes — completing X-rays, CBCT scans, and intraoral scans, billing patients directly for that work — while complex cases (full-arch reconstructions, difficult implants) are routed to accredited overseas centers of excellence with world-class implant surgeons and robotics. The initial patient cohort: Chinese-American patients in Los Angeles, the Bay Area, and New York.

The defensible value is not "cheaper treatment in China." It is:

• complete records;
• transparent total cost;
• maintainable implant systems;
• defined responsibility;
• trusted clinical handoffs;
• local follow-up;
• measurable long-term outcomes.

That list is the moat. Anyone can arrange cheap treatment abroad; almost no one can make it safe, documented, and maintainable for life.

5. The Two Pilots — Separate Tracks

Two pilots exist as parallel proving grounds, testing different halves of the vision:

The marketing pilot proves that CareFirst can deliver new patients to dental practices — the trust-building engine of Phase 1.

• Run through the "Digital Dentists" WeChat group: data-led recruitment → free audits (capped at ~20 clinics) → a paid 90-day pilot for 3–5 practices at $2,999 flat, prepaid → retrospective and conversion to a founding monthly rate.
• The pricing logic: prepayment prevents churn before GEO results land (2–4 months); the price point filters for committed practices; and a single implant patient covers the entire pilot cost — the ROI frame that makes the price easy to defend.
• Discipline rules: recruitment-post data must come from a real monitoring run on actual member practices, never invented; and hard selling never happens in the group — conversion moves to private messages and a short phone call.

The clinical pilot proves that implant identity and record sharing work with real patients.

• Target shape: roughly 20 implant patients and a small group of dentists from Jim's network, in a supervised record-sharing pilot.
• Hard safety rule: no real patient health information goes anywhere near AI development tools until the required legal agreements and controls exist.

Neither pilot replaces the other. The marketing pilot builds the provider relationships and cashflow; the clinical pilot validates the product the whole vision rests on.

6. AI's Role — The Governing Rule

The rule that keeps a healthcare AI company safe:

AI collects, organizes, extracts, translates, drafts, and flags. Deterministic software enforces safety and permissions. Licensed clinicians make clinical decisions.

A healthcare agent is never just a chatbot. Each one needs: a narrow job, approved context, permissioned tools, structured output, stopping rules, human approval, evaluations, and an audit history. In practice, every agent output falls into one of three lanes — fully automatic (safe monitoring tasks), auto-drafted but human-released (content publishing), or human-reviewed always (anything touching medical statements or patient communication).

Data is kept in strict tiers — public, internal, synthetic, identifiable health information, and final clinical authority — and identifiable patient data never crosses into development or experimentation without the proper agreements in place.

Broader compliance realities the design must keep answering to: HIPAA and Business Associate Agreements for anything touching patient data; fee-splitting and referral law (which is why the model is built on service fees, never referral fees); securities law the moment $FIRST takes any tradable form; China's data laws (PIPL, sensitive personal information, cross-border transfer) for the international loop; and clear allocation of responsibility for overseas complications and U.S. aftercare.

7. Business Model — The Candidate Stack

Who pays, in the order they're likely to pay:

1. Dentists pay first — for marketing services. This is the settled first revenue, and it doubles as wallet induction.
2. The B2B2C wallet path — a provider creates or enriches a patient's wallet, the patient claims and controls it, and CareFirst earns revenue from continuity services around that relationship. This is the most credible platform-revenue model.
3. Layered on top, over time: patient subscriptions (roughly $50–100/year has been discussed), provider SaaS, fees for second opinions and cross-border coordination, Passport verification, record-completeness review, treatment-center certification, laboratory workflows, maintenance programs, manufacturer integrations, and eventually evidence products, benchmarking, and network transaction services.

The foundation throughout: service fees for real value delivered — never referral fees, never token speculation.

8. Naming — Urgent Trademark Warnings

Research into the current names surfaced problems serious enough to change priorities. This is preliminary research, not legal advice — a trademark attorney should run a full clearance — but the headline findings are stark:

"CareFirst" — severe conflict. CareFirst BlueCross BlueShield is a $10-billion health insurer serving 3.5 million members, and "CareFirst" is their registered trademark. Worse, they operate directly in dental: they sell dental insurance plans, market them under "CareFirst Dental," and their corporate family even includes affiliates named "The Dental Network" and "First Care, Inc." A dental startup called CareFirst sits inside the exact territory where a giant incumbent holds the mark and actively sells. Adding "AI," "Tech," or "OS" doesn't fix it — the dominant word is still "CareFirst." A rename needs to be seriously evaluated before any public launch or fundraising.

"Care Wallet" — the name is taken. "CARE.WALLET" is a registered U.S. trademark (since 2020) for healthcare payment mobile apps, owned by the founder of Solve.Care — which is itself a blockchain healthcare company with its own Care.Wallet app. This is the closest possible conflict. The wallet product needs a new name or explicit legal clearance before any app-store launch.

"Global Implant Passport" — weak, and the space is occupied. "Implant passport" is a generic term: it's an established regulatory concept in Europe (implant documentation requirements), manufacturers publish "implant passport" materials, and practice-management software generates them. More importantly, a product already exists at implantpassport.app offering patient-controlled digital implant records with worldwide clinic discovery — functionally overlapping the Passport concept. This is both a naming problem and a competitor to study.

"$FIRST" — no blocking conflict found, though the "First" namespace is crowded (including that CareFirst BCBS affiliate literally named "First Care, Inc."). Since $FIRST has no decided form, clearance can wait.

Not yet searched: "CareCase," "FIRST Dental Network," "4 Levels of Dentistry," and — critically, given the cross-border model — Chinese (CNIPA) trademark clearance for every name.

9. The Coherent Sequence

The order of operations the whole vision converges on:

1. Settle the foundation — company entity, IP ownership, and the naming decision the trademark findings now force.
2. Run both pilots on their separate tracks: marketing pilot for provider trust and wallet induction; clinical pilot for implant identity and record sharing.
3. Prove one complete U.S.–overseas–U.S. continuity loop with real patients and a verified Passport.
4. Build the governed provider network and the internal evidence layer.
5. Add ordinary payments and test incentives with ordinary reward mechanics.
6. Introduce blockchain or $FIRST only where validated network utility requires them — never before.
7. Expand the CareFirst OS foundation into additional dental specialties.

The discipline that matters most at every step: never confuse the long-term vision with current capability. The vision is the destination; each phase must stand on its own.

10. Open Questions Register

The consolidated list of what still needs deciding, roughly in priority order:

Urgent (legal risk):

1. The rename question for "CareFirst" and its derivatives.
2. Rename or clearance for "Care Wallet."
3. Competitive and naming response to implantpassport.app.