MISMO Mic'd Up: Beyond the Standards

Brian Vieaux, President MISMO

MISMO Mic’d Up: Beyond the Standards brings you inside the work that quietly powers the housing finance ecosystem. Hosted by Brian Vieaux, President of MISMO, each episode features candid conversations with lenders, servicers, technology providers, regulators, and long-time MISMO volunteers who are building the data standards our industry runs on. You’ll learn: How MISMO workgroups actually operate Where standards reduce friction, risk, and cost How emerging tech (AI, POS, eClosing, consumer-permissioned data) connects to MISMO Practical ways to plug your team into the work

  1. 2d ago

    Down Payment Assistance Needs a Common Language

    Down payment assistance can help put homeownership within reach. But connecting a prospective buyer with the right program often requires lenders and loan officers to navigate information that is difficult to compare, exchange, and incorporate into their everyday workflows.In this special edition of MISMO MIC’D Up, Brian Vieaux, President of MISMO, welcomes Rob Chrane, Founder and CEO of Down Payment Resource, to discuss a proposed initiative to standardize the foundational data describing down payment assistance programs.Building on Rob’s HousingWire article, “Down payment assistance doesn’t need simplification; it needs standardization,” the conversation explores how a common data specification could help lenders, technology providers, and program administrators work together more effectively while preserving the different requirements and purposes of individual programs.Rob explains why the challenges surrounding DPA extend beyond awareness and education. Even when assistance is available, operational friction can keep it from becoming part of the borrower conversation. Loan officers need an efficient way to access and use program information, and the systems supporting them need a consistent way to exchange it.The proposed scope is deliberately focused. Rob describes a vendor-neutral foundation covering basic program information, including repayment terms, first-time homebuyer definitions, funding status, eligible property types, geographic coverage, and permitted uses of assistance. The effort would preserve proprietary eligibility algorithms, decisioning logic, and program maintenance methods while creating a more consistent foundation for information exchange.Brian and Rob discuss:Where DPA opportunities fall through the cracks. Why education matters, but cannot resolve every obstacle lenders and loan officers encounter when trying to use assistance programs.What a common data standard could accomplish. How consistently describing program characteristics could support better connections between DPA information and mortgage technology.Why an initial specification should stay focused. The importance of identifying a manageable set of data and a practical first deliverable that participants can implement.How to support borrowers earlier. Why geographic information should accommodate both a specific property and a prospective buyer who is still deciding where to purchase.How existing MISMO work could contribute. Brian discusses the potential to build on the Reference Model, Business Glossary, and lessons from the Housing Counseling Dataset.Who needs to help shape the effort. Why lenders, loan officers, LOS and POS technologists, housing finance agencies, state and municipal housing authorities, and other program administrators all bring essential expertise.The episode also explains where the proposal stands. Rob is gathering expressions of interest to support a formal MISMO work request. This is an early opportunity to help shape the initiative ahead of MISMO’s review process and a potential formal call for participation.As Rob puts it: “Lend us your name, not your calendar.”If your organization administers assistance programs, builds mortgage technology, or helps homebuyers access financing, this conversation is an invitation to contribute. Reach out to Rob or Brian to express interest and share the experience you could bring to developing a practical DPA data standard.

  2. 3d ago

    Can You Trace It Back? Building Trust in AI-Powered Mortgage Lending

    How can mortgage lenders use AI to reduce the cost of manufacturing a loan while preserving the evidence behind the work it performs? In this episode of MISMO Mic’d Up, host Brian Vieaux welcomes Steve Butler, CEO of TRUE, a MISMO member company, for a practical conversation about mortgage automation, data quality, and AI governance. Together, they explore how lenders can identify near-term operating improvements while establishing the oversight and traceability needed to support responsible adoption. Steve begins with the loan file itself, explaining how TRUE’s Mortgage Operations Service handles document classification, data extraction, validation, and condition clearing. He describes how automating these tasks can reduce repetitive work for processors and quality control teams, creating capacity for more complex issues and borrower interaction. Brian and Steve also discuss Steve’s distinction between the “coarse dial” of broader operational transformation and the “fine dial” of targeted improvements that lenders can pursue sooner. Steve shares his perspective on the potential savings from document and data automation and how lenders might use those savings to compete for borrowers. The conversation then turns to accountability. Drawing on the Massachusetts student-lending enforcement case, Brian and Steve discuss why mortgage executives should pay attention to AI developments in adjacent financial industries. Brian highlights an important distinction: automating document processing and using AI to make lending decisions involve different risks. The connection between them, however, is the data supporting the decision. To make traceability tangible, they follow an income figure from a borrower’s document into the loan origination system and consider what a lender would need to produce if that figure were challenged a year later. Steve explains TRUE’s approach to combining generative AI with deterministic validation methods, which he describes as a “chain of trust,” and discusses preserving data lineage as documents and information change. They also examine what meaningful human oversight looks like. Steve outlines TRUE’s approach to passing qualifying data through automatically while routing exceptions for review, and explains why reducing manual checking still requires executive attention to governance. Topics include: Where document and data automation can reduce friction in loan manufacturing. How targeted workflow improvements differ from multiyear transformation projects. Why AI risk should be evaluated by use case. What lenders should be able to trace from a data point back to its source. The roles of automated validation, exception review, and executive oversight. How vendors can bring governance into sales and implementation conversations. How MISMO’s FRAME can help smaller and mid-sized lenders organize their governance efforts. Brian and Steve close with a practical starting point: evaluate the data foundation, establish the current cost and time required to prepare a reliable loan file, and decide what evidence must be preserved before expanding automation. For lending executives, operations leaders, risk professionals, and technology teams, this episode offers a concrete discussion of how efficiency and accountability can advance together. Learn more about TRUE at https://true.ai and MISMO’s work on AI governance and industry standards at https://www.mismo.org.

  3. Oct 9

    Putting FRAME to Work: Two Lenders Share Their AI Governance Playbook

    Everyone is talking about AI governance. What does itactually take to build a program inside a mortgage company? In this episode of MISMO MIC’d Up, host Brian Vieauxwelcomes Nicole Taylor of Longbridge Financial and John Ashley of PRMG for acandid conversation about their firms’ AI governance journeys. Both share thedecisions, challenges, and practical lessons behind their programs—includingwhere MISMO’s FRAME helped them move forward and where the work is stillunderway. Nicole explains how Longbridge began by identifying the AItools already operating in its environment, understanding how employees wereusing them, and moving widely used tools into enterprise arrangements. She alsodraws an important distinction between Longbridge’s Center of Excellence, whereemployees share useful applications and ideas, and the governance processresponsible for evaluating risks and overseeing adoption. John describes PRMG’s experience building a governancestructure as requests for AI tools multiplied. Early proposals moved betweendepartments without a clear path to a decision. Bringing digital marketing,cybersecurity, vendor management, and legal together helped establish ownershipand move reviews forward. He also shares a practical communication lesson: anacceptable use policy needs guidance employees will actually read andunderstand. Throughout the conversation, both guests are transparentabout the difference between having an inventory and completing the work ofassessing risk. John discusses PRMG’s backlog of existing use cases requiringreview, even as the company establishes a more structured intake process fornew requests. Nicole explains how FRAME helped Longbridge recognize the need tomore deliberately document risks across both vendor products and internal AIusage. Topics include: • Finding AI already in use, including employee tools andcapabilities embedded in vendor products.• Using existing technology to discover usage and start constructiveconversations with employees.• Defining the scope of an AI governance program and prioritizing reviews.• Establishing clear responsibilities across business, technology, legal,compliance, and vendor management.• Addressing vendor contract language concerning the use of company data formodel training.• Translating governance policies into practical employee guidance.• Managing the resource demands of risk assessments and enterprise AI access.• Applying FRAME’s governance policy template and risk assessment tools to eachfirm’s needs.• Securing executive support and deciding what to tackle in the first 30 days. The guests describe how FRAME provided an accessiblefoundation for their work. For Longbridge, it offered a way to check anexisting policy draft and strengthen its approach to risk documentation. ForPRMG, it helped clarify governance responsibilities and structure theassessment process. Their experiences show how firms can build from a sharedfoundation while adapting their programs to their business models and risktolerances. A central theme runs through the discussion: a cleargovernance process can help organizations evaluate AI more efficiently, maketimely decisions, and give employees a practical path to responsible use. Whether your organization is getting started or refining aprogram already in motion, this episode offers a detailed look at the workinvolved—and concrete lessons from two leaders willing to share what they havelearned. Learn more about FRAME at mismo.org.

  4. Oct 2

    From Rate Lock to Loan Sale: Data, AI and Human Accountability

    More data. Faster technology. The same fundamental objective: protect mortgage margins from rate lock through loan sale. How can capital markets teams use standardized data and artificial intelligence to make better decisions—and maintain confidence in the outcomes? In this episode of MISMO Mic’d Up, Brian Vieaux welcomes Fuad Nasir, Vice President of Product Management for Hedging and Trading at Optimal Blue. Building on their conversation at the MISMO Fall Summit in Reston, Virginia, Brian and Fuad explore the connection between consistent data, transparent workflows and responsible AI across the mortgage capital markets life cycle. With roughly 20 years of experience in mortgage hedging, trading and analytics, Fuad shares his focus on connecting front-end pricing with hedging, loan sales and servicing values. He explains how greater visibility across those activities can help capital markets professionals understand risk, identify margin leakage and carry decisions through execution. The conversation returns to a foundational point: AI needs consistently defined data to support decisions people can trust. When systems rely on different definitions or repeated manual translations, teams spend more time reconciling information and have less confidence in the inputs driving their analysis. Brian and Fuad discuss how MISMO standards provide a common vocabulary that helps systems communicate and supports closer alignment among lenders, technology providers and investors. They also examine what responsible AI means when decisions carry real financial consequences. Fuad emphasizes explainable insights, repeatable processes, monitoring and auditability, alongside clear human accountability. Automation can help professionals identify changes and evaluate options, but teams still need the expertise and controls to challenge an output and determine when to act. Brian adds a personal reflection on using AI to summarize information, raising a broader question: how do we capture the efficiency of these tools while preserving the understanding and judgment needed to use them well? Topics include: • Connecting pricing, hedging and execution to support margin protection.• Reducing reconciliation and manual data translation across systems.• The role of MISMO standards in creating consistent context for analytics and AI.• Improving visibility into risk and performance throughout a loan’s journey.• Keeping human oversight and accountability central to automated workflows.• Avoiding overreliance on AI at the expense of professional judgment. The episode closes with a preview of the 2027 Optimal Blue Summit, scheduled for February 1–3 in Scottsdale, Arizona. Fuad shares what clients can expect, including hands-on learning, capital markets discussions, access to product teams and opportunities to help shape future technology. For capital markets leaders, mortgage executives and technology professionals, this conversation connects the promise of AI to the data quality, controls and expertise that make it useful.

  5. Sep 24

    Solving Mortgage Friction - 1 Issue at a Time

    There’s a cost to solving the same mortgage problem company by company, over and over again. It shows up in repeated data requests, time spent negotiating agreements, disconnected systems, and uncertainty about how to govern new technology. In this solo episode of MISMO Mic’d Up, MISMO President Brian Vieaux explores how the industry’s standards community brings lenders, technology providers, regulators, and other stakeholders together to address those shared challenges. Building on his recent HousingWire Daily conversation with Sarah Wheeler, Brian connects four areas of MISMO’s work to the practical decisions mortgage companies face. FRAME and practical AI governance Brian explains how FRAME helps companies organize their approach to AI through a governance policy template, an inventory, and a risk assessment tool. He discusses why that inventory needs to account for AI embedded in vendor products and the “shadow AI” employees may already be using. For companies getting started, his advice is straightforward: bring the people doing the work into the conversation, understand how AI is being used, and assign responsibility for keeping the governance process current. The Mortgage Compliance Dataset and examination efficiency The Mortgage Compliance Dataset, or MCD, emerged from a challenge shared by state regulators and lenders: exchanging loan data consistently for examinations. Brian shares how CSBS and Rocket brought different perspectives on that problem to MISMO, and how collaboration led to a common dataset. He explains the opportunity to reduce repeated formatting and data rework, while emphasizing that the benefits depend on implementation by lenders, regulators, and technology providers. Credit data and MISMO’s responsibility to stay neutral As the industry debates credit scoring models, reporting configurations, cost, and risk, MISMO’s role centers on the data needed to support those products. Brian explains why neutrality matters when bringing competitors together and how common data standards can help information move between systems with its meaning intact. eNotes and the agreements that affect adoption Digital mortgage adoption also depends on how easily trading partners can establish working relationships. Brian discusses the tri-party eNote bailee agreement and the effort to develop common starting language that could reduce repeated negotiation, support more consistent onboarding, and expand trading options for lenders originating eNotes. He encourages digital closing teams to include capital markets and warehouse partners when identifying what is slowing adoption. Throughout the episode, Brian recognizes the people contributing their time and experience to this work and connects their efforts to MISMO’s priorities of Awareness, Adoption, and Alignment. His closing challenge: choose one issue that creates friction in your business, bring it to the colleague who deals with it, and explore how your organization can use or contribute to the work underway. Learn more about MISMO’s resources and opportunities to participate at [MISMO.org](https://www.mismo.org).

  6. Sep 18

    Putting AI to Work in Mortgage: People, Data & Governance with Kastle’s Rishi Choudhary

    AI can add capacity to a mortgage operation. But what should it handle, when should a person step in, and how do you know it’s ready to represent your company? In this episode of MISMO Mic’d Up, Brian Vieaux welcomes Rishi Choudhary, Co-Founder and CEO of Kastle, for a practical conversation about AI across mortgage origination and servicing. From the first borrower inquiry to the final payment, they explore where AI agents can support teams, improve customer interactions, and help companies manage changing volumes. Rishi shares how his experience building Redfin’s mortgage marketplace shaped his view of one of the industry’s persistent challenges: adding operational capacity without continually expanding and contracting headcount. He walks through Kastle’s approach to AI support for loan officers, processing teams, and servicing operations, including lead screening, document follow-up, payment calls, and quality review. The conversation also examines where human judgment and relationships remain essential. Drawing on Kastle’s work with Carrington, Rishi explains the distinction between using autonomous agents for routine, high-volume interactions and giving employees real-time AI assistance during more complex borrower conversations. Brian pushes beyond the cost-savings discussion to explore borrower experience, service availability, multilingual support, and the performance measures lenders and servicers should consider. Responsible deployment is central to the discussion. Brian connects MISMO’s work on FRAME to the questions companies should ask before allowing an AI agent to speak or act on their behalf. Rishi outlines his approach to testing against applicable requirements, ongoing quality review, and producing results that lenders can audit. They also address a foundational issue: an AI agent depends on the accuracy, consistency, and accessibility of the data behind it. Disconnected systems and unreliable information can undermine even a promising use case, making data standards and integration critical parts of the adoption conversation. For leaders wondering where to begin, Rishi offers a practical starting point: identify a clear operational problem, focus on repeatable and well-documented work, and build from there. Tune in for a grounded discussion of what it takes to put AI to work in mortgage—and the people, data, and governance needed to support it.

  7. Sep 11

    From AI Experimentation to Execution: Building Trust with Cotality’s Jay Kingsley

    The mortgage industry is asking, “How are you using AI?” But are we starting with the right question? On this episode of MISMO MIC’d UP, host Brian Vieaux welcomes Jay Kingsley, President of Cotality’s mortgage business and a MISMO Board member, for a conversation about what it takes to turn AI’s potential into measurable business value. From reducing the time, touches, and cost of originating a mortgage to improving the borrower experience, Jay makes the case for keeping business objectives at the center of the conversation—and building the data and governance foundation to support them. Fresh from moderating a discussion with Fannie Mae’s Joyce Walsh and Freddie Mac’s Daniel Miller at the MISMO Fall Summit, Jay shares his biggest takeaways about responsible AI adoption. As mortgage companies move from experimentation to incorporating AI into everyday operations, the expectations change. Trying a tool and being impressed by its output is one thing. Making it part of an approved, repeatable business process requires trust, transparency, oversight, and the ability to explain how it is being used. Brian and Jay explore how MISMO FRAME provides a starting point for that work through governance policies, inventories of AI use cases, and risk assessments. They discuss why lenders need to understand where AI is operating across their organizations and be prepared to demonstrate that understanding to customers, business partners, and those reviewing their operations. Brian also makes an important distinction: neither Fannie Mae nor Freddie Mac has approved or endorsed FRAME. The discussion focuses on the value of its foundational components and the opportunity for broader industry adoption. The conversation then turns to the relationship between AI performance and data quality. Drawing on Cotality’s work across more than 22,000 data sources, Jay explains why reliable data must come with clear context, defined permissions, and an understanding of its origins and intended use. He shares Cotality’s emphasis on what it calls “fiduciary grade data” and describes its investment in semantic layers that help AI understand what individual data fields mean, how they can be used, and where they come from. As Jay puts it, AI can be an extremely powerful tool—but it can also be “extremely powerfully wrong.” Brian and Jay also examine how property and homeowner insights can help lenders strengthen customer acquisition, retention, and recapture. Jay discusses Cotality’s growing investment in serving loan officers, brokers, and lenders of different sizes with more targeted signals that help identify opportunities and support customer relationships. In an industry that spends billions on marketing, improving how lenders recognize and act on those opportunities can have meaningful business implications. To close, Jay reflects on his experience serving on the MISMO Board and the value of participating in the organization’s summits and working groups. Brian and Jay discuss how those forums bring industry participants together to learn, contribute, and develop shared standards—and why you do not need to arrive with all the answers to benefit from getting involved. Listen for a practical discussion about connecting AI ambition with business outcomes, building confidence in the data behind the technology, and putting governance into practice as adoption accelerates.

  8. Sep 4

    Behind the Scenes: How MISMO Built a Record-Breaking Fall Summit

    The MISMO Fall Summit broke records — and on this episode of MISMO MIC'D UP, we're taking listeners behind the scenes to find out how it all came together. Brian Vieaux is joined by Amy Moses, Vice President of Strategic Growth at MISMO, to recap a Fall Summit that welcomed 430 registrants, including approximately 350 in-person attendees, more than 70 first-time Summit participants, 20 sponsors, two major workshops, more than 25 general-session speakers, and roughly 20 MISMO workgroup meetings. Brian and Amy pull back the curtain on what it takes to orchestrate an event of this scale with MISMO's small but mighty team. From programming and speaker coordination to workshops, workgroups, newcomer orientation, networking, sponsorships and the countless details attendees never see, Amy explains why planning a MISMO Summit is essentially a year-round operation. In fact, planning for the August Fall Summit was already underway while the team was still wrapping up the June Summit in Louisville — and work on the January 2027 Winter Summit in New Orleans began before the Fall Summit had even started. The conversation also explores what makes a MISMO Summit fundamentally different from a traditional mortgage industry conference. There is no exhibit hall and the experience isn't built around moving from one presentation or product demo to another. Instead, the Summit brings together lenders, servicers, technology providers, GSEs and other industry stakeholders to actively participate in the work of creating and implementing standards that support the mortgage ecosystem. Amy walks through the rhythm of a typical Summit, beginning with Monday's MISMO workgroups and Communities of Practice, including the rapidly growing AI Community of Practice. She also discusses the newcomer orientation and MISMO's ambassador program, which pairs first-time attendees with longtime members of the MISMO community to help them navigate the Summit, make connections and get involved. Brian and Amy also revisit one of the highlights of the Fall Summit: the AI Governance Workshop centered on MISMO FRAME. The interactive format brought attendees together in smaller groups to discuss practical AI governance challenges, learn from peers and explore how FRAME can be applied within mortgage organizations. They also discuss perhaps the most unique part of the MISMO Summit — the workgroups themselves. Summit attendees can sit in on virtually any workgroup, listen, contribute ideas and learn from industry experts regardless of company, title or tenure. Those conversations span nearly every part of the mortgage lifecycle, from credit, servicing and secondary markets to architecture, compliance, title and closing, eMortgage and emerging technologies. The episode also highlights MISMO Boot Camp and the path toward the Associate MISMO Standards Professional (AMSP) and Certified MISMO Standards Professional (CMSP) designations. Brian shares another development: MISMO certification activities will provide opportunities to earn points toward the Mortgage Bankers Association's Certified Mortgage Banker (CMB) designation, creating another connection between MISMO education, industry engagement and professional development. Finally, Amy gives listeners a preview of what's coming next. Following a successful Fall Summit workshop focused on standardizing the Tri-Party eNote Bailee Agreement, MISMO plans to formally advance that effort into a workgroup, with its first in-person meeting expected at the January Winter Summit. Whether you've attended dozens of MISMO Summits or have wondered what actually happens at one, this episode offers a look at the people, planning and collaboration behind the experience — and why more mortgage professionals are discovering that MISMO's work touches their businesses whether they realize it or not.

About

MISMO Mic’d Up: Beyond the Standards brings you inside the work that quietly powers the housing finance ecosystem. Hosted by Brian Vieaux, President of MISMO, each episode features candid conversations with lenders, servicers, technology providers, regulators, and long-time MISMO volunteers who are building the data standards our industry runs on. You’ll learn: How MISMO workgroups actually operate Where standards reduce friction, risk, and cost How emerging tech (AI, POS, eClosing, consumer-permissioned data) connects to MISMO Practical ways to plug your team into the work