PollyOS vertically integrates the full capital markets lifecycle on one data foundation, with every basis point attributed from rate sheet through settlement

SAN FRANCISCO: #alongcamepolly--Polly, the only AI-native, vertically integrated mortgage capital markets operating system, today introduced PollyOS, completed by the launch of Polly Hedge. With next-gen Hedge joining Polly's AI-native Product, Pricing, and Eligibility (PPE) engine, and the reintroduction of Loan Trading Exchange, every loan is now priced, hedged, and sold on a single engine, from a single record, with every basis point of margin attributed from rate lock through settlement. PollyOS is the only platform to unify the full capital markets lifecycle on one data foundation.

PollyOS was purpose-built as one foundational operating system on Polly's AI-native architecture and patented agentic framework. This distinction matters. Historically, capital markets incumbents were assembled through product acquisitions, each carrying a different architecture and infrastructure, vocabularies, data models, and versions of the truth. No integration layer can fully close those seams.

"The industry has never had a real answer to a simple question: 'How much money did we actually make on that loan?' Every capital markets leader knows that the gap between target and realized margin exists, and until now the best anyone could do was guess it after the fact," said Adam Carmel, Founder and CEO of Polly. "PollyOS solves this critical legacy problem with precision, loan by loan, as it happens. That depth of loan-level attribution, in real time and throughout the loan's lifecycle, is only possible when pricing, hedging, and trading are never separate in the first place."

ONE ENGINE, ONE LOAN RECORD

With PollyOS, a rate sheet, lock, hedge position, and loan sale are not multiple events across multiple systems. They are one loan record, priced by one engine, and updated in real time to account for any market movement or changes to the loan. When a lock comes in, the position moves. When a loan changes status, cancels, or funds intraday, the pipeline reflects it in that moment rather than in the next morning's file. When the market moves, the desk sees it against live pricing rather than a prior-day snapshot.

Nothing is exported, imported, or needs to be reconciled; the position on the screen is the position the lender holds at that moment.

The same architecture aligns the secondary desk and accounting. PollyOS expresses gain and loss using the calculations accounting already relies on, layered with hedge economics. The month-to-date profitability the C-suite sees mid-month is the same number that ties to GAAP reporting at close, so answers to margin erosion questions are no longer stitched together from disparate systems or delayed until month end. Now lenders can make informed business decisions in real time and optimize revenue on every single loan.

Jon Foy, SVP of Product at Polly, shared: "Like everything we do, PollyOS was built in close and continuous collaboration with our customer partners. We know legacy tools have historically handicapped lenders' ability to identify margin erosion and make informed business decisions, and that is a problem Polly's R&D team was determined to solve. Delivering the industry's first true capital markets operating system is a milestone our team has worked toward for years, and the most exciting part is what that makes possible from here."

EVERY BASIS POINT EXPLAINED

Polly Hedge is the key unlock that completes PollyOS. An AI-native pipeline risk management and hedge analytics system, Hedge was built into the core of PollyOS rather than connected to it, drawing on live lock data, forward execution data, real-time LOS updates, and timely market data. Polly spent more than two years building this transformational approach around a single requirement: that every PollyOS function deliver measurable value and return to mortgage industry participants.

"A hedge desk runs on judgment and trader knowledge that nobody writes down. It lives with one or two people, and when they are out, the lender's risk posture goes out with them," said Brandon Story, SVP of Strategy at Polly and former Chief Investment Officer at Mr. Cooper. "PollyOS makes that knowledge structural. Every basis point of margin is explained the same way for every loan, every day, by the system rather than by the one person who knows where to look."

BUILT WITH THE INDUSTRY, FOR THE INDUSTRY

Polly developed Hedge in close collaboration with capital markets leaders and advisors who have spent their careers building and running hedge programs, including Gary Malis of Paramount Residential Mortgage Group Inc. (PRMG) and Rob Kessel, Founder and former CEO of Compass Analytics.

"One of the most complex challenges in mortgage banking is accurately attributing profitability at the individual-loan level. You are connecting a margin established on day one with an execution weeks later, with market movement, hedge performance, pull-through, loan-level economics, and investor demand all occurring in between," said Gary Malis, Chief Strategy and Capital Markets Officer and Partner at PRMG. "What I find compelling about PollyOS is the granularity they have brought to that attribution problem. The value is not simply knowing what you made or lost on a loan, it is understanding why and using that information to make better decisions going forward. That creates valuable information not just for the capital markets team, but for the entire boardroom. Just as importantly, Polly's willingness to collaborate with existing partners and incorporate them into the broader scope of the platform makes this an especially exciting rollout."

Rob Kessel, Founder of Panoramic Capital Academy, served as an advisor to Polly throughout the development of Hedge. "I have spent three decades in mortgage capital markets, much of it building the systems lenders still use today. The industry has always talked about leakage in gross, estimated terms because that was the best that legacy architecture and workflow allowed. Polly started from a different premise entirely: one integrated engine, one loan record, and loan-level attribution as a property of the operating system rather than a report generated after the fact. That premise changes what a capital markets team can know, as well as how fast they can act on it."

The initial cohort of Polly customer partners adopting PollyOS has been onboarded, with broader availability following MBA Annual.

"PollyOS is what we set out to build when we founded the company," Carmel concluded. "Our customer partners now run their entire capital markets operation on one engine, with a level of precision the industry has only ever imagined. And because the foundation is AI-native, what that engine can do will continue to expand rapidly. We are just scratching the surface of what is possible, and this is a major milestone for our team to accelerate our vision."

Polly will attend the 2026 MBA Annual Convention and Expo in Chicago, October 11-14. #MBAAnnual26 attendees are encouraged to schedule a meeting with Polly's attending team to see PollyOS in action and run loan-level attribution on their own volume and margin numbers.

ABOUT POLLY

Polly is the only AI-native, vertically integrated mortgage capital markets operating system, engineered to maximize lender profitability through precision margin optimization, agentic automation, and real-time, loan-level attribution and profitability analysis. By unifying product, pricing, and eligibility (PPE), pipeline hedging, loan trading, and secondary accounting on a single AI-native data foundation, PollyOS establishes a definitive source of truth that aligns key stakeholders around consistent forecasts, traceable gain-on-sale economics, and margin strategies grounded in live loan performance. Continuous machine learning and agentic workflow automation empower the nation’s leading banks, credit unions, and independent mortgage lenders to translate secondary market execution into bottom-line P&L impact. To learn more, visit www.polly.io or follow Polly on LinkedIn.

Fonte: Business Wire