Payerset launches Research Assistant for healthcare price transparency

Payerset issued a Business Wire announcing Research Assistant, an AI-powered answer engine built on more than 20 trillion records of contracted rates processed each quarter, unifying negotiated rate data, claims, policy, and market benchmarking.

HEALTH desk — a hospital that wants to know whether its joint-replacement rates match the rest of its market has been handed payer files too big to open. An answer engine that names the facilities, the codes, and the Medicare benchmark before it answers is how a contracting team gets that comparison without hiring a data team.

The line above the lede, and what the lede says the launch is. The FinancialContent page prints this sentence above the lede: “Jiro Health Among the First Health Tech Companies to Extend Payerset’s Intelligence from Health Systems to Individual Clinicians and Practices.” This desk did not see that sentence on a businesswire.com page. The lede says Payerset, the infrastructure company behind healthcare price transparency, announced the launch of Research Assistant, a fully self-service, AI-powered answer engine built on more than 20 trillion records of contracted rates processed each quarter, unifying payers’ own negotiated rate data, claims, policy, and market benchmarking into a single platform. Hospitals and health systems now have complete market context, including both the most current view of the market and a full, longitudinal record of how rates have shifted over time, along with real, plain-language answers, without needing an analyst, a data team, or specialized training. More than 20 trillion is the release’s scale. Twenty trillion is 20,000,000,000,000. That expansion is this desk’s. The count is the release’s. Longitudinal, here, means the history of the rates, not only today’s file. Self-service, here, means a person can use it without a services project. Those glosses are this desk’s. The release prints the words. Each quarter is the lede’s clock. A later about sentence says per quarter. They are the same claim, worded twice. Do not file them as two different counts.

Why the release says the files were unusable, and whose history that is. In 2022, when compliance for the Transparency in Coverage rule took effect, the federal government required health plans to publish their negotiated rates, with the intention of giving hospitals, employers, and patients a clear window into what healthcare actually costs. The result, the release says, was a mountain of machine-readable files, shortened to MRFs, that no one could actually use: a single payer’s pricing file can contain billions of rows, often published in inconsistent formats and riddled with errors. A machine-readable file, here, is the computer file a health plan posts, not a PDF a person can skim. That gloss is this desk’s. The release spells out MRF. Billions of rows is the release’s description of one payer file. It is not the 20 trillion figure. Do not collapse them. Payerset was founded that same year to solve that problem, building the infrastructure required to normalize the data and render it usable across the formats healthcare organizations rely on. 2022 is the release’s year for the rule’s compliance start and for the founding. This desk did not open a payer file, and it did not read the rule text for this filing.

Jerry DiMaso, as a quote, not as a study. Jerry DiMaso, CEO of Payerset, said that for years hospitals and health systems have had access to pricing data without ever really having access to the insights that mattered most to each of them, like how their rates stack up against the market. He said they’d get a file so massive that no one on their team could actually make sense of it. He said that instead of handing someone a spreadsheet with more data, the company is fixing healthcare pricing by delivering the right answer, instantly, in the hands of the people negotiating on behalf of patients. Instantly is his adverb. It is not a stopwatch this desk held. Fixing healthcare pricing is his sentence. It is not a measured price change. Those are his sentences on the release. This desk did not interview him.

Jiro Health, as the release states the partnership, and Greg Field as a quote. The release says Payerset’s price transparency data powers some of the largest and most innovative healthtech solutions in the country. Jiro Health is a Practice Intelligence platform that turns real-world data into personalized clinical, operational and financial insights for U.S. clinicians and physician-led practices. The companies’ work together enables Jiro to integrate Payerset’s normalized pricing and claims intelligence directly into its platform. That market intelligence, the release says, gives clinicians and practices actionable, data-driven insights for reimbursement benchmarking, RVU calculation, denial remediation, contract negotiations and growth planning. RVU, a relative value unit, is a unit Medicare uses to value a service. That gloss is this desk’s. The release prints the letters and does not expand them. Greg Field, founder and chief executive of Jiro Health, said health systems have struggled to get real market visibility into this kind of data, and for individual clinicians and small practices it has been even further out of reach. He said that with the partnership, Jiro can now deliver actionable financial intelligence for individual clinicians they can’t see anywhere else. Some of the largest, and can’t see anywhere else, are the page’s words. This desk did not log into Jiro, and it did not count the clinicians.

What Research Assistant builds on, and the two ways in, on the release. The launch builds on Payerset’s existing products: Rate Explorer, which the page prints with a trademark mark, and which lets managed care and finance teams compare negotiated rates directly, and Data Lake, also trademarked on the page, which gives health tech companies and consultants API access to that same data. An API, an application programming interface, is a way for one piece of software to ask another for data. That gloss is this desk’s. Research Assistant, the release says, is the company’s next evolution: a full intelligence platform available in two forms. A Model Context Protocol server, shortened to MCP, lets healthcare organizations plug Payerset’s pricing, claims, and policy data directly into the enterprise AI tools they already use, including Claude and ChatGPT, without needing a separate interface. A companion portal product embeds that same intelligence in a chat-style experience within Payerset’s own platform, for teams that want direct, conversational access without their own AI integration. Including Claude and ChatGPT is the wire’s list. The wire body this desk read does not name Microsoft 365 Copilot or Gemini Enterprise. Do not add those two names to the wire. MCP, here, is a shared plug so a chat tool can call another company’s data. That gloss is this desk’s. The release prints the words. This desk did not connect a chat account.

The questions the release says a person can ask, and Jacob Little as a quote. In both formats, the release says, users can move beyond raw market data and ask direct, strategic questions, instantly receiving a data-backed answer, whether that is how their negotiated rates compare to peer health systems in their market, which states offer the strongest reimbursement for a given service line, or where a fee schedule gap is costing them revenue before they ever sit down at the negotiating table. A fee schedule, here, is the list of what a payer pays for each service. That gloss is this desk’s. Instantly, again, is the release’s adverb. Jacob Little, chief commercial officer of Payerset, said the company built this tool to solve a real, specific problem, and the market’s response has told them just how badly it was needed. He said health systems that would normally take months to evaluate a new product are telling them to just turn it on. He said that is the clearest signal they could ask for that they have built something the industry needed, even before it had a name for it. Months, and just turn it on, are his sentences. They are not a customer count or a timed sales cycle this desk measured. Signal, in that sentence, is his word for a sign from buyers. It is not this desk’s channel. Those are his sentences on the release. This desk did not interview him.

Who the about box says the company is, and what the two older products do, still on the release. Founded in 2022 by chief executive Jerry DiMaso and chief commercial officer Jacob Little, Payerset has grown without institutional capital, remaining profitable since early in the company’s history. It processes more than 20 trillion records of contracted rates per quarter and serves hospitals, health systems, consulting firms, health tech vendors, and other organizations working to make sense of healthcare pricing data. Without institutional capital, and profitable, are the about box’s claims. This desk did not read a cap table or a tax return. The about box calls Payerset the infrastructure layer behind healthcare price transparency, ingesting, normalizing, and delivering every payer machine-readable file in the country so hospitals, health systems, consultants, and health tech companies can use pricing data without building a data engineering team. Rate Explorer, the box says, is a fully self-service platform that gives managed care, finance, and contracting teams instant, no-code access to compare negotiated rates across payers, providers, and billing codes, model fee schedule scenarios, and benchmark performance against the broader market, with data refreshed quarterly and full historical data always retained. Data Lake is an enterprise-grade API that lets health tech companies, analytics platforms, and consulting firms embed pre-normalized pricing and claims intelligence into their own products, unified across every payer and updated on the same ongoing basis. Quarterly is the box’s refresh clock. No-code means a person can use the screen without writing a program. That gloss is this desk’s. Every payer file in the country is the box’s claim. This desk did not count the files. The media contact on the page is SolComms, payerset@solcomms.co. This desk did not email that address.

What the card shows, and what the release does not print. The card is the Research Assistant research-plan screen from the product page at payerset.com/platform/ai-research-assistant/, a headless crop. The question on the still is “How do our joint replacement rates compare to every other health system in our market?” The screen then says Your health system, resolved. The counts on the still are 6 facilities, 9 TINs, 1,284 NPIs, and 5 peer systems. The page markup stores those as data-count values and prints 0 until a script fills them. The still shows the filled counts. The research guide is marked generated. Scope: Your 2 hospitals, 5 peer systems, 1 market. Codes: MS-DRG 469, 470, and CPT 27447, 27130. Benchmark: Medicare IPPS, current fiscal year. Validate: Cross-file rate match, unit basis. The foot line is “Plan ready. Starting research.” The screen prints 6 facilities and also “Your 2 hospitals.” This desk does not collapse those into one number. A page bullet says the product maps every facility, TIN, and NPI, including affiliates that file under other names. That bullet is the page’s. It is not a sentence that 6 and 2 are the same count. TIN, a taxpayer identification number, is the tax id a provider bills under. NPI, a national provider identifier, is the id for an individual clinician or an organization. MS-DRG is a Medicare Severity Diagnosis-Related Group, a code for a hospital stay. CPT is a procedure code. IPPS is Medicare’s inpatient payment system. Those glosses are this desk’s. The screen prints the letters. The numbers on the still are the page’s demo. They are not a hospital this desk audited. The wire’s schema image is the Payerset logo. This filing does not use that logo. The card has no date text.

What the product page and the MCP page add, and what they are not. The product page says Research Assistant is the next evolution of Rate Explorer. It runs inside the Payerset platform on the same data, and the self-service filtering remains. The difference, the page says, is where you start: with the question, rather than with the filters. The page’s questions-and-answers say the same research tooling is available through Payerset MCP, a hosted server live in Claude, ChatGPT, Microsoft 365 Copilot, and Gemini Enterprise. Those four names are the product page’s and the MCP page’s. The 24 Sep wire names Claude and ChatGPT and says including. It does not print the four-name list in the body this desk read. The MCP page says Payerset MCP is the first healthcare price transparency MCP server live in all four major AI platforms, with customers in production on each, and that the same hosted endpoint serves all four. First, and customers in production, are that page’s claims. The page does not name those customers. The endpoint printed on the page is mcp.tpi.payerset.com/mcp. The MCP page also prints “Quadrillions of records, full rate history.” Quadrillions is that page’s scale line. It is not the wire’s “more than 20 trillion records of contracted rates processed each quarter.” Do not collapse them. Access, the MCP page says, uses per-user OAuth, data is retrieved live rather than indexed into the AI platform, and entitlements are enforced on the server on every call. OAuth, here, is a sign-in that lets one app act for a user without handing over the password. That gloss is this desk’s. This desk did not connect an assistant, and it did not read a customer contract.

Plain English for the rest of the card. Research Assistant is the answer engine. Rate Explorer is the click-and-filter tool it builds on. Data Lake is the API for other companies. An MRF is a payer’s machine-readable price file. More than 20 trillion records each quarter is the wire’s scale for contracted rates. Billions of rows is the wire’s description of one payer file. Quadrillions is the MCP page’s separate scale line. 08:00 AM EDT is 8:00 a.m. Eastern and 12:00 p.m. UTC. 6 facilities, 9 TINs, 1,284 NPIs, and 5 peer systems are the research-plan screen’s demo counts. Your 2 hospitals, 5 peer systems, 1 market is the research guide’s scope line. MS-DRG 469 and 470, and CPT 27447 and 27130, are the codes on that guide. Medicare IPPS, current fiscal year, is the benchmark line. Jerry DiMaso is chief executive. Jacob Little is chief commercial officer. Greg Field is the Jiro Health quote. Claude and ChatGPT are the names on the wire. Copilot and Gemini Enterprise are the extra names on the product and MCP pages. This filing is the 24 Sep announcement.

PRIMARY here: Payerset’s 24 Sep 2026 Business Wire, release 20260924176038, stamped September 24, 2026 at 08:00 AM EDT on the FinancialContent syndication this desk read — Tier A PRIMARY, the company’s own announcement. businesswire.com itself did not load. The launch of Research Assistant, the more-than-20-trillion sentence, the negotiated-rate, claims, policy, and benchmarking sentence, the no-analyst and no-data-team sentence, the 2022 Transparency in Coverage paragraph, the billions-of-rows sentence, the DiMaso quote, the Jiro paragraph and the Field quote, Rate Explorer and Data Lake, the MCP sentence naming Claude and ChatGPT, the companion portal sentence, the three example questions, the Little quote, the about box including no institutional capital and profitable, the quarterly refresh, and payerset@solcomms.co are that release’s. The four platform names, the quadrillions line, the endpoint, and the research-plan counts are the product page and the MCP page. They are not sentences this desk read on the wire body. NOT claimed: that this desk opened businesswire.com, that first-of-its-kind was audited, a named hospital behind the 6 facilities, that 6 facilities and 2 hospitals are the same count, that 20 trillion or quadrillions was independently counted, a named MCP customer, a customer count behind “just turn it on,” a dollar price, that this desk connected Claude or audited the $31,920 demo line on a later panel of the product page, a stock tip, or investment advice. The card is the research-plan UI. Distinct from the already-filed trilliant-health-oria, verana-health-agent-claire, veeva-study-builder-agent, and bighat-biosciences-series-c.