Press & media — CareCost
Last updated: June 12, 2026
There’s a moment, right after hope, when a patient asks, “how much will it cost?” — and watches your face for the answer.
CareCost exists so the people across the desk can answer. We parse the negotiated-rate files insurers are required to publish — public in name only — into pricing intelligence for the people fighting to make medicine affordable.
The story
In specialty pharmacy, one question decides everything: what will this infusion actually cost this patient? Answering it honestly takes about 45 minutes of hard math — per patient, per drug, per plan. The drug’s price lives in one public file, the payment rules in another, the assistance programs on a manufacturer page that changed last quarter. Every number exists. Nothing connects them. So the people who do this work for a living build their own workarounds — notes, bookmarks, formulas — because the tool that should exist, didn’t.
In March 2026, founder Erin Rose started building it. The software turned out to be the easy part — because the data it needed, public by federal mandate, was public in name only.
Under the federal Transparency in Coverage rule, insurers must publish every negotiated rate in machine-readable files. They comply the way you’d expect: files the size of small libraries, formats that change payer to payer, indexes that point nowhere. CareCost parsed them anyway — 60 terabytes of source files across the largest national payers, 151 terabytes cataloged across 37, distilled into a normalized database of more than 36 million negotiated rates covering 353,000 billing codes in all 56 states and territories — re-imported and re-verified every quarter.
What came out the other side stopped being a calculator. One drug. One dose. Forty prices — the same medicine, the same code, and negotiated prices that differ several-fold depending on who’s paying. That’s not a pricing system anyone can see; that’s a battlefield, and almost nobody fighting on it has the map. CareCost is the map: cost estimation for specialty practices (Estimate), channel-economics intelligence for infusion programs (Optimizer), the full negotiated-rate corpus, queryable (Explorer), and the normalized data itself, delivered programmatically (API).
CareCost, LLC is founder-operated, based in Portland, Oregon, and early — paid subscriptions are live and the company works hands-on with design-partner practices. It takes no money from payers, manufacturers, or PBMs.
Fact sheet
- Legal entity
- CareCost, LLC (Oregon)
- Founded
- 2026
- Founder
- Erin Rose
- Headquarters
- Portland, Oregon
- Products
- CareCost Estimate · CareCost Optimizer · CareCost Explorer · CareCost API
- Audience
- Specialty practices, infusion programs, billers and patient-access teams, researchers
- Data scale
- 60 TB parsed / 151 TB cataloged · 36M+ normalized rates · 353K billing codes · 56 states & territories
- Independence
- No funding or compensation from payers, manufacturers, or PBMs
- Press contact
- editorial@carecostestimate.com
Stories this data can support
We’ll run the queries for any of these — see “Working with our data” below.
- The price of the same thing. The same procedure code, the same market, and negotiated rates that differ several-fold between payers. We can produce the spread for any billing code, payer set, or region in the corpus.
- Transparency in name only. What it actually took to make “machine-readable” insurer files readable: 151 terabytes cataloged, formats that differ by payer, indexes pointing to files that don’t exist. A story about the gap between complying with a rule and honoring it.
- What the fiduciary can finally see. Employers sign off on health plans without seeing the rates their own plan negotiated. Those rates are public now — and checkable.
- One founder and an AI (for technology and business press). CareCost has no data team. One founder plus AI collaborators parsed 60 terabytes of hostile file formats — a working example of what a single domain-disciplined human and current AI tooling can actually ship.
Citable numbers
All figures as of June 12, 2026, verified by query against our production systems and refreshed quarterly. Methodology: /methodology.
What we collect
What’s queryable today
The reference layer
Public context layers
We also load the public CMS registries — 9 million+ providers (NPPES) and 6,101 hospitals (HCRIS cost reports) — to join rates to real-world entities. These are public files, listed for completeness, not claimed as proprietary.
Custom statistics — rate-spread distributions, payer coverage gaps, negotiated-vs-public comparisons — available on request for working journalists.
The products
The real out-of-pocket cost of a specialty drug — assistance programs included — in seconds, not 45 minutes.
Channel economics for infusion programs: margins, white-bag exposure, and reimbursement intelligence drawn from the source files.
Every negotiated rate, queryable. The map of what healthcare actually costs, payer by payer.
CareCost’s normalized rate corpus, delivered programmatically.
Founder
Erin Rose — Founder, CareCost LLC
In specialty pharmacy, one question decides everything — what will this infusion actually cost this patient? — and answering it honestly takes 45 minutes of hard math, because every number is public and nothing connects them. Erin Rose founded CareCost to build the tool that should have existed. She parsed the 60 terabytes of “machine-readable” rate files insurers publish but no human can read, and built the products that put real prices in practices’ hands. Before CareCost, she spent twenty years building AI-augmented data and revenue systems inside national media organizations — including leading AI adoption for a 200-person sales organization. She lives in Portland, Oregon.
Speaking topics:
- Transparency in name only — what it takes to read the data insurers “publish.”
- One person + AI: parsing 60 terabytes without a data team.
- The price-tag moment — why drug affordability fails at the front desk, not the pharmacy.
- Building AI-augmented operations inside organizations that didn’t ask for them.
For interviews: editorial@carecostestimate.com
Headshot (web resolution): download →
Working with our data
Working on a pricing story? We’ll run the query. CareCost provides journalists and academic researchers with custom data pulls and comped access to CareCost Explorer — the full negotiated-rate corpus — in exchange for attribution. Qualification is simple: a working story or research question, and a byline or institution we can verify. Start with an email describing what you’re chasing: editorial@carecostestimate.com.
Questions we get asked
How did you get this data — is that legal?
It’s public data. Federal rules (Transparency in Coverage, 45 CFR Part 147) require insurers to publish every negotiated rate in machine-readable files. We download what the law makes public and do the engineering nobody else wanted to do.
Do you handle patient data?
No. CareCost processes payer rate files and public assistance-program rules — no patient data, no PHI, ever. We compute prices, not people.
Who funds you?
CareCost is founder-funded and early: paid subscriptions are live, and we work hands-on with design-partner practices. What it is not: funded by payers, manufacturers, or PBMs — no investment, no sponsorship, no placement fees. The full independence statement is in our Editorial Policy.
Why should anyone trust a company this small with numbers this big?
Every figure on this page traces to a public source file you can re-derive yourself; the methodology is published; corrections run on a public log. We’re glad to walk any journalist through the pipeline end to end — source file to printed number.
Copy-paste boilerplate
For use in articles, panel introductions, or filings. Cite as “CareCost” (company) or full product names on first reference.
One-liner
50-word boilerplate
120-word boilerplate (standard)
120-word variant (for feature/profile pieces — emotive open; offer when a writer wants color)
Brand assets
The CareCost mark, for editorial use. Please don’t alter the artwork or use it to imply endorsement that hasn’t been granted.
Everything in one download — logo files, headshot, fact sheet, and usage note: carecost-brand-kit.zip.
Editorial standards
CareCost publishes a full Editorial Policy covering source hierarchy, AI use, conflict-of-interest rules, correction handling, and review cadence. Each reference page carries a “Last reviewed” date and is re-reviewed at minimum annually even if no source has changed.
Corrections: corrections log or editorial@carecostestimate.com. Confirmed corrections normally ship within five business days; corrections that affect a billing or coding decision in current use are prioritized.
Press contact
Erin Rose, Founder — interview requests, data pulls, methodology questions, corrections: editorial@carecostestimate.com. One-page fact sheet: download (PDF).