amazon
Prepared by
Joe Nalley
show your work.

Independent healthcare programs
for the largest private workforce in America.

Claims verification. Behavioral health navigation. Substance use treatment. Maternity steering. Each with performance guarantees, measurable outcomes, and a 90-day pilot designed to prove value on your own claims data.
joe.nalley@showyourwork.health · joe-nalley.com
Confidential · Prepared exclusively for Amazon · Not for distribution · June 2026joe.nalley@showyourwork.health
Joe Nalley
02
Your Workforce

1.1 million U.S. employees. The majority
do physical labor. The injury data is public.

3x
Amazon warehouse injury rate vs. Walmart in 2023
Strategic Organizing Center 2024 report
50%+
Of all recordable warehouse injuries are musculoskeletal disorders
OSHA 300 logs / SOC analysis 2024
~700K
Warehouse and fulfillment workers — lifting, packing, sorting shifts up to 12 hours
Amazon 10-K 2024 / press reports
6.6
Serious injuries per 100 warehouse workers in 2023 — nearly double the industry average of 3.5
Bureau of Labor Statistics / SOC 2024
$1.1B+
Estimated annual behavioral health + SUD spend at industry benchmarks for a population this size
Milliman / SAMHSA 2024 employer benchmarks
CVG
Air Hub in Hebron, KY — one of the largest cargo facilities on Earth. Thousands of associates in our operating geography.
Amazon press / KY Economic Development
More than half of Amazon's recordable warehouse injuries are MSK. The clinical pathway from back injury to opioid prescription to dependency doesn't require modeling — it's the most documented occupational health pipeline in America.
All figures sourced from public filings, OSHA logs, and industry benchmarks. We verify against your claims data during pilot.02 / 12
Joe Nalley
03
The Gap

You built a pharmacy. You launched a $5 plan.
The structural gaps are still open.

Haven failed. The problem it tried to solve didn't.

Amazon, JPMorgan, and Berkshire Hathaway spent three years and shut down Haven in 2021. The thesis was right — employer healthcare needs independent governance. The execution tried to build everything from scratch instead of deploying proven, narrow programs against specific cost categories.

The $5 plan is a cost lever, not an outcomes lever

The 2026 basic plan reduces Amazon's per-employee cost. It doesn't reduce readmissions. It doesn't verify billing. It doesn't navigate a warehouse worker through SUD treatment. Plan design optimizes premiums. Program design optimizes outcomes.

Amazon Pharmacy fills prescriptions, not gaps

Pharmacy integration and the RxAdvance PBM switch optimize drug cost. But behavioral health navigation, SUD treatment, maternity steering, and claims verification aren't pharmacy problems. They're clinical governance problems. No PBM solves them.

Claims verification at 1.1M employees

Self-insured at this scale means high-cost claims flow daily. Your TPA processes them. Your TPA audits them. Nobody independent verifies before payment. At a 4% discrepancy rate on claims above $25K, the math is tens of millions annually in preventable overpayment.

You test everything — from delivery routes to warehouse ergonomics to packaging dimensions. Your healthcare benefit spend is one of the few cost categories that doesn't get the same treatment.
03 / 12
Joe Nalley
04
The Portfolio

Four programs. Each independent.
Each designed to prove itself in 90 days.

Lead Program

Caliber

Pre-payment billing verification — advisory-only claims governance
Verifies high-cost claims (>$25K) before payment using a seven-check Billing Governance Standard. Issues a Billing Governance Certificate. Advisory-only — never denies claims, routes findings to your team. Independent of your TPA. At 1.1M employees, even a fractional improvement yields eight figures.
4–6% expected discrepancy rate · 10 business day turnaround · verifythebill.com
Lead Program

Curated

Independent behavioral health navigation with readmission warranty
One named navigator per member across every BH episode — acute psych, IOP, PHP, residential SUD, eating disorders, PTSD. 13 clinical areas. Quality-scored provider database across 13 clinical areas. 30-day readmission warranty on SUD residential and acute psych. Built for the warehouse workforce profile: burnout, injury-related depression, and the SUD pipeline your injury rate predicts.
200K+ patients managed · 80%+ residential completion · Single-digit readmission · curatedhealth.care
Supporting Program

Continuum

Virtual-first OUD/SUD treatment with buprenorphine and navigator continuity
Virtual-first MAT program. Buprenorphine via permanent DEA telehealth pathway. Named navigator manages the full episode from intake through stabilization through step-down. KY and OH operations align directly with your CVG Air Hub and fulfillment center geography. 42 CFR Part 2 compliant.
9/week IOP standard · 42 CFR Part 2 compliant · Fee-first, earning into risk · continuumhealth.care
Supporting Program

Waybright

Warranty-backed maternity navigation for self-funded employers
Claims-integrated early risk identification at weeks 16–20. Facility-level quality data in the consumer's hands. Delivery site steering. 30-day readmission warranty held by credentialed delivery providers. Our platform fee is at risk against maternity outcomes across your population. At 1.1M employees, maternity is a material cost category every year.
$1,677 modeled savings per navigated pregnancy · 4-tier steerability · waybright.health
Supporting Program

Cadence

Specialty therapy continuation governance — published clinical framework
At 1.1M employees, specialty pharmacy spend is enormous. Cadence evaluates whether specialty therapies should still be running. 65,234 patients across 3 locked cohorts including NIH federal validation. 29.1% flag rate. Published in Pharmacy Times (April 2026). CGS v1.1 — 8-section framework, 7 reassessment triggers, 4 drug classes.
29.1% flag rate · $14.3M first-cycle savings · $3–6 PEPM · showyourwork.health
Each program operates independently. Adopt one, two, or all five. No bundling requirement. No platform lock-in.
04 / 12
Joe Nalley
05
Caliber · Claims Verification

At 1.1 million employees, a 4% billing
discrepancy rate isn't a rounding error.

Seven-check Billing Governance Standard

Every claim above $25K runs through CODEVAL (procedure coding accuracy), DUPCHECK (duplicate detection), RATEVAL (rate verification), DURVAL (duration validation), CREDVAL (credential verification), SVCVAL (service verification), BUNDLEVAL (unbundling detection). Seven checks. One certificate.

Advisory-only. Never denies.

Caliber issues a Billing Governance Certificate. It never touches the claims adjudication process. Findings route to your benefits team with full documentation. You decide what to act on. We provide the intelligence. You keep the authority.

Independent of your TPA

Your TPA processes your claims. Your TPA audits your claims. That's self-grading. Caliber is structurally independent. No downstream economic interest in volume. No relationship with your network providers. No reason to miss anything.

The Math at Amazon's Scale
Amazon is self-insured for approximately 750K+ covered lives (employees and dependents). High-cost claims above $25K — inpatient stays, surgical episodes, NICU admissions, BH residential — flow through your TPA daily.

At a conservative 4% billing discrepancy rate, a self-insured plan this size typically sees $20–50M+ in preventable overpayments annually. Caliber's PEPM cost is a fraction of a single recovered claim.

This isn't about finding fraud. It's about finding billing errors, unbundling violations, duplicate charges, and rate discrepancies that happen at scale in every health system — and that nobody catches when the auditor is the same entity processing the claim.
4–6%
Expected discrepancy rate on high-cost claims
OIG / AHLA / MGMA benchmark data
10 days
Standard turnaround. 5 days for short-hold TPAs.
$3–5
PEPM tiered by population
750K+ covered lives = deepest tier
For Amazon: Caliber is the highest-ROI entry point. Fastest proof, lowest integration burden, largest dollar recovery at your scale.05 / 12
Joe Nalley
06
Curated · Behavioral Health

700,000 warehouse workers.
The behavioral health load is predictable.

Your warehouse workforce faces a specific clinical profile: physically demanding shifts, high injury rates, burnout, isolation from traditional support systems, and limited access to behavioral health beyond an EAP. The injury-to-depression pipeline and the burnout-to-substance-use pipeline are the two most predictable BH pathways in employer health — and both run through your fulfillment centers every day.
The Model
01
One named navigator per member across every BH episode. Not a call center. Same person from intake through discharge through follow-up.
02
13 clinical areas — acute psych, residential SUD, IOP, PHP, eating disorders, PTSD, adolescent BH, dual diagnosis, detox, crisis stabilization, outpatient, neuropsych, co-occurring medical.
03
30-day readmission warranty on SUD residential and acute psych. Active after 100 measured episodes. If a navigated member is readmitted for a preventable condition, the credentialed provider absorbs the cost under their delivery warranty — and our platform fee is forfeited for that episode.
04
Quality-scored provider database. We don't deliver care. We don't own facilities. No downstream economic interest in where your employee goes.
Why Amazon

Your EAP offers a handful of sessions. That's triage. It's not treatment for a warehouse associate dealing with chronic pain, depression from a workplace injury, or a substance use disorder that started with an opioid prescription after a back injury on the pick line.

Curated fills the gap between what the EAP starts and what your associates actually need — longitudinal navigation across multiple episodes, with a named navigator who stays through every step. At 1.1M employees, even modest BH engagement generates thousands of navigated episodes per year.

200K+
Patients managed across BH episodes
80%+
Residential completion rate — vs. 50% national average
SAMHSA TEDS 2023
06 / 12
Joe Nalley
07
Supporting Programs

Two more programs. Same pilot structure.
Same 90-day proof window.

Continuum · Substance Use

The injury happened at work. The opioid prescription happened after.

Virtual-first MAT program with buprenorphine via permanent DEA telehealth pathway. Named navigator from intake through stabilization through step-down. Your warehouse demographics — physically demanding labor, high injury rates, rural and semi-rural fulfillment center locations — match the SUD risk profile precisely. KY and OH operations align directly with the CVG Air Hub and surrounding fulfillment centers.

$9,000+
Average employer cost per SUD episode without early intervention
SAMHSA / Milliman 2024
67%
Of opioid-dependent workers cite workplace injury as origin
NSDUH 2023 / NIOSH
42 CFR Part 2 compliant · Fee-first model · continuumhealth.care
Waybright · Maternity

The most expensive predictable episode.

At 1.1M employees plus dependents, Amazon sees thousands of pregnancies annually. Claims-integrated risk identification catches complications 4 weeks earlier than standard screening. Facility-level quality data — C-section rates, NICU admissions, readmissions by hospital — in her hands. Delivery site steering toward high-quality, lower-cost facilities.

$20,416
Average commercial maternity episode
Peterson-KFF 2024
$71,158
Average NICU admission. 1 in 10 pregnancies.
HCCI 2021
30-day readmission warranty · $4.50–8.50 PEPM · waybright.health
Continuum and Waybright are independent of Caliber and Curated. Adopt any combination. Each proves itself on your data.
07 / 12
Joe Nalley
08
The Pilot

You A/B test everything else.
We're asking for the same discipline on healthcare.

Amazon's operating culture is built on test-measure-scale. You don't roll out a fulfillment workflow to 1,000 buildings without proving it in 10. We don't ask you to commit to a multi-year contract without proving the numbers on your own claims data. The pilot is the product — 90 days, your baseline, your scorecard. Performance guarantees and warranties are active from day one.

Scope

Defined population. Minimum 1,000 lives for BH/SUD programs. Claims verification scales to your full book from day one. You choose the segment — a single fulfillment region, a specific TPA book, a benefits tier. We prove on the population you pick.

Timeline

90-day measurement window. Claims-verified outcomes, not self-reported surveys. Baseline established from your prior 12 months. Outcomes measured against your own history, not an industry average.

Guarantees

Performance guarantees are active from pilot day one. Our platform fee is at risk from the first navigated member. If population outcomes miss the baseline, we forfeit the at-risk portion.

Exit

If it doesn't work, you walk. No termination fee. No penalty. No lock-in. You keep the data, the baseline analysis, and the scorecard. The 90 days were free intelligence either way.

The Scorecard
Every pilot produces a claims-verified scorecard at day 30, 60, and 90. Metrics are agreed before launch — not after we see the results. You define success. We measure against it. The scorecard is the decision document for full deployment.

Caliber: discrepancy rate, prevented overpayment, turnaround time, certificate accuracy
Curated: readmission rate, completion rate, cost per episode, navigation engagement
Continuum: treatment retention, MAT adherence, ED utilization, return-to-work
Cadence: flag rate, review influence rate, continuation appropriateness, savings per cycle
Waybright: C-section rate, NICU rate, readmission rate, steering acceptance
Why This Works for Amazon
You already think this way. You test warehouse layouts, delivery algorithms, packaging materials, and Prime pricing models before committing at scale. Haven failed partly because it tried to build a platform instead of testing narrow programs against specific cost categories. We're the narrow test.

Three things you learn in 90 days that you can't learn any other way:
1. Whether your claims baseline matches what you think it is
2. Whether navigation changes employee behavior, or just generates reports
3. Whether a vendor will put their own fee at risk against their own performance
The pilot isn't the beginning of a sales cycle. It's the end of one. Either the numbers move or they don't. That's the whole conversation.
Pilot pricing: PEPM at standard rates. No setup fees. No implementation charges. Warranties and guarantees active from day one.08 / 12
Joe Nalley
09
Unit Economics

What the 90 days cost.
What we expect them to return.

ProgramPEPMAnnual CostPrimary Lever
Caliber$3–5$1.8–4.5MPrevented overpayment
Curated$3–7$1.8–6.3MReadmission + completion
ContinuumPer patientVariableED diversion + retention
Waybright$4.50–8.50$3.6–7.7MNICU reduction + steering
Cadence$3–6$2.7–5.4MContinuation appropriateness
Illustrative: ~750K covered lives (employees + dependents)
Pilot: start with one program, one population segment
PEPM tiers by volume — 1.1M employees = deepest tier pricing
We publish the economics honestly. If modeled savings don't exceed program cost in the expected case, we say so. The remainder buys guarantees and risk transfer.
Caliber ROI Signal
At a 4% discrepancy rate on claims >$25K, a 750K-life self-insured plan typically sees $20–50M+ in preventable overpayments annually. Caliber's PEPM is a rounding error against the recovery. At Amazon's scale, a single quarter of Caliber findings can exceed the full annual program cost by an order of magnitude.
Curated ROI Signal
Average BH readmission costs $12,000–$18,000. Curated's readmission warranty means every prevented readmission is pure savings. At 1.1M employees, BH episode volume generates hundreds of navigated cases per quarter — single-digit readmission rates vs. the 15–20% industry norm move real money.
Portfolio Discount
Two or more programs: 15–20% combined PEPM discount. Three or more: 20–25%. Single scorecard, single data feed, one vendor relationship across all programs.
All economics verified against your claims data during pilot. We don't model in a vacuum.09 / 12
Joe Nalley
10
Competitive Position

What we do that your current
infrastructure structurally can't.

Dimension
Amazon Pharmacy
EAP / Existing BH
Haven (Failed)
Show Your Work
BH depth
Not BH-specific
Limited sessions. Referral.
Never built
13 clinical areas. Full episode chain.
SUD treatment
Fills prescriptions
Rehab referral only
Never built
Virtual MAT. Buprenorphine. Named navigator.
Maternity
Not maternity-specific
Not maternity-specific
Never built
Claims-integrated. Facility steering. Warranty.
Claims audit
None
None
Conceptual only
Pre-payment. 7-check BGS. Advisory-only.
Cost guarantee
Drug cost savings
Engagement metrics
None shipped
Provider-held delivery warranty. Platform fee at risk. The only vendor in the category that warrants outcomes.
Pilot terms
Internal build
Annual contract
JV. Shut down after 3 years.
90 days. Walk-away. No penalty.

We don't replace Amazon Pharmacy or your existing benefits infrastructure. We fill the gaps they weren't designed to fill — with programs that prove themselves on your claims data before you commit.

10 / 12
Joe Nalley
11
The Team

Built from the payer seat.
Designed for the employer buyer.

Joe Nalley
Joe Nalley
Founder & CEO

15 years building healthcare organizations and governance structures. Specialty risk products at a Fortune 25 payer — maternity, MSK, oncology, CHF — across millions of commercial members. Direct visibility into how payers price, manage, and lose money on the exact episodes these programs address.

Built and scaled a 13-location health system (BH, SUD, MAT, primary care, surgical center, community hospital). 30,000+ patients. Exited. Founded ClearBill — $9.2M returned to payers in the first six months. Exited. The programs in this deck aren't concepts. They're built on the same architecture, sold to the same buyer, governed by the same standards.

Cadence: 65,234 patients · 3 independent cohorts · NIH federal validation · Published in Pharmacy Times
M.S. Applied Behavioral Analysis, Northeastern · joe-nalley.com
Platform
7 Products Built and Published
Curated, Continuum, Waybright, Caliber, Cadence, Compass, Covenant — each with its own site, canon, and governance standard. Same infrastructure. Same employer buyer.
Payer Fluency
Speaks Your TPA's Language
Claims integration, risk stratification, actuarial reserving, and stop-loss coordination are built in — not outsourced to a consultant after the sale.
Capital
Founder-Led, Capital-Backed
Active capital conversations with healthcare-focused investors. Employer pilots are the proof points. Every signed pilot validates the model and accelerates deployment.
11 / 12
Joe Nalley
12
Next Steps

One conversation. One pilot.
90 days of your own data.

Haven proved the thesis and failed on execution. The $5 plan proves the cost pressure.
The window for narrow, provable programs is open right now.

Week 1–2

Introductory call. Identify which programs fit. Agree on pilot population and scorecard metrics.

Week 3–4

Claims data review. Baseline analysis. Pilot agreement signed. No setup fees.

Day 1–90

Programs live. Scorecards at day 30, 60, 90. Warranties and guarantees active throughout.

Day 91

Decision. Scorecard in hand. Continue, expand, or walk away. No penalty either way.

Timing
Pilot can launch within 30 days of agreement. No implementation timeline. No technology integration required for initial deployment. Claims data feed is the only prerequisite. A one-page pilot term sheet is available on request.
Amazon's Moment
Haven failed in 2021. The $5 basic plan launched for 2026. Amazon Pharmacy is scaling. The RxAdvance PBM switch is underway. Every move signals the same thing: Amazon is actively rethinking how it manages healthcare cost. The pieces that are still missing — behavioral health navigation, claims verification, SUD treatment, maternity steering — are exactly what we built. The timing isn't coincidental. It's structural.
Start Here
We recommend starting with Caliber (claims verification) and Curated (behavioral health) — highest ROI density at your scale, fastest proof, lowest integration burden. Add Continuum and Waybright in wave two once the scorecard proves the model.
joe.nalley@showyourwork.health · joe-nalley.com · showyourwork.health
Confidential · Prepared exclusively for Amazon · Not for distribution · June 202612 / 12