A dozen federal datasets, one searchable platform — and an AI assistant that speaks plain English.
Sanjay M. Udoshi MD
The data to understand American healthcare prices is public — it is just scattered across a dozen federal datasets, none of which talk to each other. MediCosts pulls them into one place: a single searchable platform for cost, quality, and payment transparency, with an AI assistant that answers questions across every dataset in plain English.
MediCosts aggregates more than 48 million records across 20+ CMS and federal sources. The hospital explorer covers 5,400+ facilities with DRG pricing set against HCAHPS ratings, readmission and safety penalties, and star scores — all sortable, and now drillable into a single-hospital profile. The landing analysis frames the stakes immediately: average markup over Medicare, penalized hospitals, and safety failures, with the worst price-gougers and readmission penalties surfaced on arrival.
Two of the most revealing datasets are now first-class: Open Payments, with 30M+ Sunshine Act payments from pharmaceutical and device makers to physicians, searchable down to the individual provider; and Medicare Part D, with five years of drug spend across 14,000 drugs at prescriber-level detail. A geographic view maps cost, quality, and access to the county level, overlaying CMS spend with HRSA shortage areas and CDC community-health signals.
Abby, the MediCosts assistant, streams natural-language answers across all of these datasets, so a question like “which hospitals near me charge the most for a joint replacement, and how do their ratings compare?” returns an answer instead of a spreadsheet. Access is now gated by single sign-on through Authentik, consistent with the rest of the platform fleet.
The next surface is consumer-facing: a “My Healthcare Costs” experience that turns the same public data into a personal, before-the-bill estimate — the difference between knowing what hospitals charge in the aggregate and knowing what your care is likely to cost.