The biomarker-native financial planning platform for the decumulation decades.

Executive frame

SCQ · situation-complication-question-answer.

Governing thought

The one sentence that governs every downstream decision.

HIFP wins by owning the biomarker-native planning decision at the point of trust — the concierge physician — with an actuarially grounded planning agent at the user surface, and a published Advice-Boundary Architecture no competitor can replicate at speed.

Three consequences follow:

1. The MVP is dual-surface consumer, AI-native at the user layer, actuarial under the hood. Anchor deals with 1–2 national concierge chains (MDVIP, One Medical Premium (via Amazon), Executive Health Group, PartnerMD) put HIFP inside a captive HNW patient base; a four-tier D2C freemium funnel runs in parallel from launch; the Planning Agent + Plan-Delta monitor ship at MVP as the flagship demo, not as H2 features.
2. The moat is model-class + regulatory architecture, not distribution. HIFP publishes stratified back-tests on shorter-horizon decision-specific claims (LTC-onset window, cognitive-decline threshold, high-cost care-episode probability) where population life tables are weakest — and publishes the Advice-Boundary Architecture whitepaper as an auditable framework. Two compounding IP assets, neither transpilable at speed.
3. API/OEM, institutional data-product, and employer-executive-benefit are Y2–Y3 monetization variants, documented and pre-architected but not the raise story. Variants A (OEM), B (institutional), and D (B2B2E) activate only after the consumer business is self-sustaining. Contingency Variant C (B2B-only pivot) documented as fallback.

Locked positioning

Full pitch: HIFP is the biomarker-native financial planning platform for the decumulation decades — physician-distributed, actuarially grounded, and delivered through a personalized planning agent gated by a published Advice-Boundary Architecture no competitor can replicate at speed.

Elevator (≤15 words): The biomarker-native planning platform for the decumulation decades — physician-distributed, actuarially grounded.

Naming discipline (holds in every artifact)

• "Planning Agent" — not "Copilot," never "agent" alone. Descriptive, defensible under advice-boundary scrutiny.
• "Agentic" — reserved for exactly two things: the Plan-Delta monitor and internal service ops. Everything else is "AI-assisted" or "AI-grounded." Prevents overclaim.
• "AI-native" — used only at the user surface. The actuarial model is not AI; it is actuarial. Overclaim fails technical DD.

Problem space

Three data points quantify the market; three challenges + three risks survive the v3 tightening.

Problem statement. HNW individuals aged 45–70 make multi-decade financial decisions using health inputs no more sophisticated than a single life-expectancy number, and no existing product turns biomarker, wearable, or clinical data into the individualized planning parameters those decisions actually require. The trust to collect and act on that data lives with physicians, not advisors — specifically with the concierge practices already serving this segment — but concierge practices have no financial-decision surface, no infrastructure to build one, and no AI capability disciplined enough to ship a consumer-facing decision-support agent inside health-finance regulatory boundaries.

Three converging inflections that did not exist in 2024

1. The dollar gap. Fidelity's 2026 estimate: $185,500 per 65-year-old, $345,000 per couple, before LTC. Americans 55+ hold $124T (73%) of US household wealth.
2. Data + AI availability. Function Health at 100K+ paying members, $2.5B Series B. Frontier LLMs now capable of bounded tool-use + structured-output patterns required for a regulated consumer agent.
3. The trust asymmetry. Concierge medicine serves ~300–500K HNW patients — longitudinal biomarker data, existing consent, willingness to pay $2K–$25K/yr cash-pay medicine.

Three risks that survive the v3 tightening

1. Executional — Fatima's transition timing is the biggest single go/no-go signal.
2. Regulatory — WA MHMDA + IL GIPA still apply; novel-jurisprudence risk mitigated via Planning Agent geo-gating at MVP.
3. Category-timing — anchor value in the Fidelity $185K number + concrete decision surfaces, not longevity or AI category-heat.

Product strategy

The strategic thesis in three sentences.

Health is the largest un-modeled variable in a 30-year plan, and it has just become measurable at consumer scale. Frontier AI has, in the same window, become capable enough to ship a bounded consumer decision-support agent — but only for teams disciplined enough to survive the regulatory boundaries where health, finance, and AI intersect. HIFP wins by owning the biomarker-informed planning decision at the point of trust — a concierge physician for the HNW patient, and HIFP itself for the direct consumer — with a decision-grade actuarial model published to scrutiny, a planning agent as the flagship consumer experience, and a published Advice-Boundary Architecture competitors need meaningful time to replicate.

Four product pillars

AI folded into each pillar as connective tissue, not appended as a fifth.

Pillar 1 · Biomarker-Native Risk Model + AI-Assisted Model Development

Bet. Decision-grade CI narrowing on shorter-horizon, biomarker-sensitive events (LTC-onset window, cognitive-decline threshold, high-cost care-episode probability, personalized SWR). Applied ML accelerates model iteration + drift detection; underlying actuarial model remains defensible under ASA-LTC scrutiny — no neural-net-in-the-loop for LTC-onset until Y2 at earliest.

Guardrail. Model outputs are planning parameters, never user-facing clinical predictions. Enforced by the Advice-Boundary Classifier at runtime.

Now (MVP). ASA-LTC-led model trained on Kaiser NorCal 5-year LTC cohort + Framingham cognitive derivatives + concierge-partner longitudinal panels. Two published stratified back-tests.

Pillar 2 · Planning Translation Layer + Planning Agent (flagship)

Bet. Users want "budget $340K for care with these year-by-year draws", not "life expectancy 87.3 with 95% CI". The Planning Agent turns Monte Carlo output into that language — bounded, structured, deterministic-templated, classifier-gated.

Guardrail. Translation is versioned, documented, template-based. LLM never invents numbers; it selects and sequences model output.

Now (MVP). Planning Agent Q&A mode + Plan-Delta monitor ship at MVP for Plan Core and above. Free tier gets scoped preview (5 Q&A interactions/mo). Planning Agent geo-gated to non-WA + non-IL residents at MVP; H1 expansion after production evidence.

Next (H1). Brief mode (advisor + physician) · Roadmap mode · native mobile · physician cohort brief generator · advisor share-brief per-platform templates.

Pillar 3 · Consumer Surface — 4 tiers, 2 fronts (identity)

Pillar 4 · Trust-First + Advice-Boundary Architecture — six commitments

1. Individual data is personally identifiable to the user only.
2. We never sell your data. Ever. Aggregated or individual.
3. Model training uses HIFP user data ONLY where the user has opted in to training-set inclusion.
4. Physician-facing dashboard aggregated at minimum cohort size of 25.
5. Export or delete data at any time; deletion rights transfer to acquirer as contractual term.
6. Every AI-generated output is bounded tool-use, versioned template, classifier-gated, audit-logged. Full architecture publicly documented in the Advice-Boundary Architecture whitepaper (principles at H1; full architecture at Series B).

Strategic variants

Documented but not the MVP raise story.

Solution

One sentence, three benefits, four KPI categories.

HIFP is a biomarker-informed planning platform delivered to HNW individuals across four consumer tiers. It ingests biomarker, wearable, and self-report data through AI-assisted multi-modal extraction; produces individualized distributions on four named decision surfaces (safe-withdrawal-rate, LTC sizing, housing-timing, care-reserve) via an actuarially grounded model; delivers those outputs through a bounded planning agent (Q&A mode at MVP, geo-gated to non-WA + non-IL); watches user data continuously through a Plan-Delta monitor that proactively re-runs plans and notifies on material trajectory changes; publishes a curated fee-free directory; and lets users share plan outputs with their existing advisor or physician on a per-share, user-controlled basis.

Three benefits (rule of three, quantified)

1. Reserves sized to the individual, not the average. Target: ≥20% CI narrowing on care-reserve estimate for users with concierge-source biomarker data vs. Fidelity-anchored baseline.
2. Life decisions timed to trajectory. Target: 80%+ of MVP users report using at least one HIFP output in a household financial decision within 90 days.
3. Physician-endorsed acquisition, AI-safe by construction. Target: sub-$100/patient CAC in concierge channel; Advice-Boundary Architecture the physician's compliance officer can read and approve.

Use cases

Three consumer-facing use cases at MVP + variant use cases documented for Y2+.

1 · Concierge-embedded LTC-timing decision, Planning-Agent-mediated

62-year-old MDVIP patient opens HIFP in the MDVIP portal. HIFP ingests her lab feed (multi-modal extraction from Function panel PDF) + Oura data. She asks the Planning Agent: "How does my April Function panel change my LTC purchase decision?" Agent runs bounded scenario, classifier passes, output: care-reserve plan showing 72% probability of moderate-care event years 8–11, reserve $310K–$440K, specific LTC purchase-timing window. Two weeks later Oura HRV deviates; Plan-Delta monitor re-runs plan, pushes notification. Her MDVIP physician sees aggregated cohort dashboard (no individual outputs unless she shares).

2 · D2C safe-withdrawal-rate for a Function Health member

58-year-old, $4M investable, pays $39/mo Plan Premium. HIFP ingests Function panel + Oura. Personalized SWR range 4.6–5.1% vs. 3.7% flat-line default. She engages the Planning Agent: "What happens if I retire at 62 vs. 65?" — sees side-by-side scenarios. Increases annual draw $50K without materially raising ruin risk. Exports AI-generated advisor brief to her existing eMoney-using RIA.

3 · Physician-endorsed household decision (H1 with household pricing)

68-year-old couple in an MDVIP practice under household plan. Joint mobility + cognition-decline modeling → housing-timing plan (downsize year 3 vs. remodel + age-in-place through year 10). MDVIP care team sees aggregate dashboard + AI-generated pre-visit brief highlighting the couple's approaching housing-timing inflection. MDVIP pitches HIFP as differentiating benefit at renewal.

Target markets

Two primary MVP channels; one prioritized Y2 optionality; three Y2–Y3 variant markets.

• Concierge chains (MVP anchor). MDVIP, One Medical Premium (via Amazon), Executive Health Group, PartnerMD. Fatima leads exec-to-exec BD; sales cycle 6–12 months.
• D2C HNW segment (MVP parallel). ~5M households with $1M–$25M investable; Function/Superpower/HLI overlap; freemium funnel with concierge co-marketing.
• B2B2E employer executive-benefit (Y2 prioritized). F500 CHRO + benefits-consultant channel (Mercer, Willis Towers Watson, Aon).
• Variant markets (Y2–Y3): wealth-platform OEM (Variant A); insurer/reinsurer data-product (Variant B); B2B-only pivot (Variant C contingency).

Team

Founder + two prospective founding executives. Each closes three DD questions.

HIFP is Dr. Sindhu Pandit's founding concept. Fatima and Brandon are prospective founding executives — both pending. Fatima is pending Gate 1 full-time commitment; Brandon is pending founder-agreement close. The v3 raise-materials framing is honest about this pending status.

Dr. Sindhu Pandit — Founder (title TBD — COO / CMO / Chief Clinical Officer, open for reconciliation)

MD (physiatrist), MBA SMU. Currently VP, Clinical Product Strategy at Hyro (conversational AI in healthcare); formerly Clinical Leader on Salesforce Global H&LS. Concept originator; founding equity holder; clinical anchor.

Closes three DD questions:

1. Clinical concept validity. MD physiatrist who conceived the biomarker-informed planning thesis end-to-end; not a founder outsourcing clinical to a hire.
2. Physician-community access. Sindhu's Hyro + prior Salesforce H&LS clinical relationships + Fatima's current Oracle Health + prior Salesforce + Microsoft + Allscripts + McKinsey network + boutique-practice pipeline + clinical-conference visibility gives the founding team a real bench for practice-owner conversations and adversarial-test-set clinical adjudication.
3. Concept-to-clinical-validation execution ownership. Active author on stratified back-tests + clinical adjudication for the adversarial-test-set adversarial classes.

Dr. Fatima Paruk — Chief Executive Officer (prospective — pending Gate 1)

Currently in a senior health-leadership role at Oracle (LinkedIn stale as of 2026-09-02; title confirmation pending direct input). Formerly SVP & Chief Health Officer, Salesforce H&LS; McKinsey Clinical Center of Excellence lead; Microsoft H&LS CMIO; Allscripts CMO.

Closes three DD questions:

1. Health-strategic distribution. Current Oracle Health + former SVP & CHO Salesforce H&LS + former Microsoft H&LS CMIO + former Allscripts CMO network reaches MDVIP, One Medical Premium (via Amazon), Executive Health Group, PartnerMD at the buyer level directly — the concierge-chain BD wedge is not cold outreach.
2. CEO gravitas + investor room. McKinsey Clinical CoE lead + physician-executive credibility opens the health/longevity-strategic investor lane with warm intros, not the seed circuit.
3. F500 CHRO relationships + regulatory strategy. Direct networks into Mercer / Willis Towers Watson / Aon executive-benefits divisions + regulatory-counsel relationships (Manatt / Perkins Coie) unblock Variant D activation and Tier 1 gap #4.

Brandon Stauber — Chief Product / Technology Officer (prospective — pending founder agreement)

Two-exit founder-operator (iNetEvents 1999→2004; The Wine Spies 2007→2012). Dual Salesforce Certified Architect (System + Application) + Salesforce AI Associate. Currently Director, Partner Innovation Engineering at Salesforce — creator of the Agentforce Partner Innovation Lab. Prior: Salesforce Health Cloud + EHR/EMR + telemedicine at Saxa (acquired by J2 Interactive); FinTech partner-lending platform at Direct Capital / CIT Bank. 25+ years across health, fintech, consumer eCommerce. Role split trigger: combined CPO/CTO through Series A → CPO after VP Eng hire in Y2.

Closes three DD questions:

1. Regulated-industry AI product execution. Agentforce Partner Innovation Lab creator + dual SF Certified Architect + SF AI Associate — the specific discipline HIFP needs to ship the Planning Agent + Advice-Boundary Architecture.
2. Health Cloud + EHR/EMR + regulated-data-integration precedent. Saxa Solutions — architected Salesforce Health Cloud programs + EHR/EMR integration + native iOS/Android telemedicine for medical device + specialty providers.
3. FinTech partner-ecosystem + platform-integration precedent. Direct Capital / CIT Bank — owned partner-lending platform end-to-end (marketing impression → GL), partner-facing origination portal, B2B APIs.

MVP scope

12 months, ~30% scope-cut from v3 draft per Red Team v3 T2. Sindhu aligned on smaller-MVP posture.

Ships at MVP

• 1 concierge chain live · anchor MSA + patient-portal integration
• 4 consumer tiers · freemium + Core + Premium + concierge-embedded
• Actuarial risk model · 2 published stratified back-tests (LTC-onset + cognitive-decline)
• Planning Agent Q&A mode · geo-gated to non-WA + non-IL residents
• Plan-Delta monitor · proactive replanning + notifications
• Advice-Boundary Classifier · release-gated at 0% zero-tolerance failures + ≥98% bounded-tolerance
• Multi-modal onboarding ingestion · target < 5 min to first plan
• SOC 2 Type II · in flight, certified within 12 months of MVP
• Salesforce Sales Cloud · internal CRM only (Y2 upgrade path from HubSpot bridge)

Deferred to H1

• Planning Agent Brief mode (advisor + physician) · Roadmap mode
• Native iOS + Android · ship on PWA at MVP
• Genomic ingestion (Galleri, HLI) · separate IL GIPA consent flow
• Curated fee-free directory · content curation ramp
• Share-with-incumbent AI-generated export · advisor-platform templates
• Multi-provider LLM · Anthropic single-provider at MVP
• Physician cohort brief generator · post chain 1 live
• Advice-Boundary Architecture principles post (full whitepaper at Series B / Q3 2028)

Commercial model

Five revenue streams, sequenced across four horizons.

1. Concierge per-patient license (MVP · primary) — $10–15/patient/mo, chain-paid.
2. D2C freemium + Plan Core + Plan Premium (MVP · primary) — blended $22/mo paid ARPU.
3. Variant D · B2B2E employer (Y2 · prioritized) — $200–500/exec/yr.
4. Variant A · Consumption-only OEM (Y2 · activation-gated) — $50K–$300K ACV/platform.
5. Variant B · Institutional data-product (Y3 · activation-gated) — 7–8 figure model-license deals.

Raise structure

$7–9M Series A led by a health/longevity-strategic investor (a16z Bio + Health, ARCH Venture, GV, Lux Capital, health-strategic vehicles), with AI-forward generalist lane opened in v3 (Sequoia AI-application, Radical Ventures, Bond, Lightspeed AI) as legitimate second-lane. Target close Q3 2027.

Technical summary

Six-service Python + TypeScript modular monolith on AWS + Cloudflare + managed infra.

• Consumer surfaces: Next.js + PWA (MVP); React Native (H1); chain patient-portal iframe integration.
• Backend: Python 3.12 (FastAPI) for core + model layer; TypeScript (Hono) at the edge.
• Data: Postgres 16 (RDS) + TimescaleDB for wearable time-series + pgvector for RAG; S3 for object storage; Snowflake for warehouse (H1+).
• LLM: Anthropic Claude single-provider at MVP; OpenAI + Google added in H1 for classifier voting + fallback.
• Auth: Clerk (consumer) + WorkOS (enterprise SSO for concierge chain + B2B2E).
• Payments: Stripe (Billing + Tax).
• Internal CRM: Salesforce Sales Cloud earns internal-CRM slot; HubSpot as MVP bridge, SF migration when 5+ AEs hired. SF Data Cloud deferred to Y3 buyer-agnostic decision. Agentforce rejected for consumer Planning Agent (per-user licensing kills freemium; lose Advice-Boundary Architecture control).
• Compliance + observability: Vanta or Drata (SOC 2); Datadog + Sentry + Arize/WhyLabs (model observability).

Ten-move Advice-Boundary Architecture

Structured intent + template rendering · bounded tool-use (~15 named functions) · two-layer safety net · deterministic numeric base · adversarial red-team test set as release gate · counsel sign-off per release · full session audit log · user attestation on first session · jurisdiction-aware consent flow (WA + IL enhanced + geo-gated at MVP) · model-vendor pass-through terms + AUP compliance evidence log. Detailed in AI Narrative v3.

Confidence + assumptions

Confidence 0.64. Seven assumptions. Each falsifiable.

Seven assumptions

1. National concierge chain signs anchor MSA within 90 days of active outreach. Gate 2
2. Freemium free-to-paid CVR ≥6% @ 90d, ≥12% @ 180d, at <$3/free-user/yr infra. Gate 4
3. ASA-LTC-led back-test shows ≥20% CI narrowing on LTC-onset window. Gate 3
4. Fatima commits full-time within 6 months. Gate 1
5. Regulatory architecture achievable at MVP within 25% engineering-cost inflation. Gate 5
6. Chain-level per-patient economics work at $10–15/mo. Gate 6
7. Advice-Boundary Classifier holds at 0% failure on zero-tolerance classes (free-text financial recommendation, diagnostic claim, consent violation, cross-user leak) AND ≥98% on bounded-tolerance classes. Gate 7 — reworded per Red Team v3 T8. Fallback if fails: v2-shape MVP.

Key gaps

Ranked by impact-to-raise. 21-item merged punch list; top tiers shown below.

Tier 1 · Existential

1. Concierge-chain anchor MSA · Fatima closer; Sindhu clinical; Brandon integration architecture.
2. ASA-LTC actuarial lead hire.
3. Fatima full-time commitment.
4. Dual-jurisdiction regulatory counsel (Manatt / Perkins Coie / DWT / Cooley).
5. Adversarial red-team test set curation partner (in-house vs. Anthropic red-team services / HiddenLayer / Robust Intelligence).

Tier 2 · Deck-blocking

Longitudinal-cohort dataset access (Kaiser NorCal + Framingham derivatives) · D2C conversion + CAC pilot · Founder-employment structure documented · Concierge economics validation (3+ chain-exec conversations) · Multi-provider LLM contract terms.

Pro-forma financials

Base case; downside case ~$100–300M exit; upside case $700M–$2B exit.

Four bidder pools

What changed vs. v2

Executive delta for founder reconciliation.

1. Top-line locked — biomarker-native + decumulation decades + physician-distributed + Advice-Boundary Architecture as moat.
2. AI is first-class product identity, not implementation detail. Planning Agent + Plan-Delta + published architecture.
3. "Agentic" disciplined — reserved for Plan-Delta monitor + internal service ops only.
4. Pillar 4 gets a sixth commitment — every AI-generated output bounded, template-rendered, classifier-gated, audit-logged.
5. Fourth bidder pool — AI-native strategic added (directional).
6. Brandon founder profile corrected — Chief Product / Technology Officer (prospective, pending founder agreement) with combined title through Series A. Not "co-founder" — HIFP is Sindhu's founding concept and equity anchor.
7. No Y1 CPO/CTO hire — savings redirected; staff+ Founding Engineer #1 as safety net.
8. Investor room widens — AI-forward generalist lane added beside health/longevity-strategic.
9. 6-service Python + TS modular monolith locked as MVP architecture.
10. Salesforce earns its place once (Sales Cloud, Y2 upgrade path); Agentforce rejected; Data Cloud deferred.
11. MVP scope tightened ~30% per Red Team v3 T2; Sindhu aligned.
12. Planning Agent geo-gated to non-WA + non-IL at MVP; H1 expansion after production evidence.
13. 7 assumptions with reworked Gate 7 (two-part failure-class structure per Red Team v3 T8).
14. Confidence 0.62 → 0.64 (down from v3-draft's overclaimed 0.68 per Red Team v3 T7).