FICO announced the release of two foundation models. FICO Focused Language mainly deals with conversations and the language aspect of finance to determine fraud and process documentation for loans. On the other hand, FICO Focused Sequence works best for transaction analytics. Trust Score is a key component of what makes the two FICO models effective for heavily regulated industries, such as finance. The Trust Score serves as a guardrail that indicates how closely a response aligns with its training data. The Trust Score also takes into account context found in the data. So if the model is used to read through documentation about European financial instruments, the Trust Score can see if the response is relevant. A response with a high score indicates that it is accurate in terms of its data coverage and is not misleading. Responses with low scores may prompt the bank to review its data or refine how the model responds. FICO FLM works best on understanding the language used in transactions. It has two general use cases. The first is for compliance and communications. It understands the rules governing how financial institutions can and should provide information to customers and extract information from conversations. What is special about FLM is that since it monitors the back and forth between a bank and a person, it can detect if the customer is undergoing some financial hardship. The bank can tailor its approach to providing information to them, taking into account their economic position. The second use of FLM involves underwriting, which is the act of offering a loan or capital to an individual or a business. The model can take into account the person’s interactions with the bank and review loan documentation. FICO FSM deals with transaction data. “The architecture is different; it has something called a contrastive head and a supervised head,” Scott Zoldi, chief analytics officer at FICO said. “The contrastive head says, is this transaction in or out of pattern, while the supervised head says, is this change in behavior fraud or not. The supervised task knows the probability of fraud, the fact that she has hardship, and we have to intervene.”
Bank of America, Wells Fargo and U.S. Bank now offer predictive checking account insights and forecasts based on transaction patterns, helping customers manage finances and potentially encouraging them to consolidate banking relationships
According to Keynova, three major banks — Bank of America, Wells Fargo and U.S. Bank — have started offering predictive consumer checking account insights within their mobile apps. Using factors like past transaction patterns, scheduled and recurring transactions and expected deposits, their apps give consumers a sense of the road ahead. Bank of America’s Erica virtual assistant presents insights upfront after the mobile user logs in. This includes such information as payment due dates and recurring subscription charges. Insights are also presented about spending habits using the categorized spending and cash flow over the previous year. Wells Fargo flags its forecast on the app’s pay and transfer screen (point 1 in the first screen below). Tapping on the activity forecast icon (point 2) sends the user to further detail, along with a timeline (point 3). A link (point 4) on the same screen leads to a screen (point 5) presenting any scheduled transactions, along with a chart forecasting balances in the future. U.S. Bank’s app presents a balance forecast on the app’s account details screen. In a general sense, it gives a forecast of whether the balance over the next 30 days will be positive or negative (point 1) and then the bank’s estimate of what the running balance will be (point 2). A popup box (point 3) gives the user the data elements that went into the bank’s forecast. Susan Foulds, managing director of Keynova, believes these features can encourage consumers to consolidate more of their relationships in a single institution, because the more that’s under the umbrella, the better the forecasts can be. She says other banks that aren’t as far along as Bank of America, Wells and U.S. Bank could be doing more even with data they already have. Even simply totaling up whatever is scheduled for the next week to 14 days, and presenting the math, is a service to customers.
Snapdocs integrates eClosing platform with Vesta LOS enabling complete closing process management for 25% of U.S. real estate transactions directly in Vesta’s LOS
Digital closing provider Snapdocs announced a new integration between its eClosing platform and Vesta, a next-generation loan origination system (LOS) and a provider of mortgage origination software. With Snapdocs powering one in four U.S. real estate transactions, this integration lets mutual customers manage the full closing process directly within the Vesta LOS. It’s designed to streamline workflows, cut manual work and speed closings from origination to final signature. Mike Yu, co-founder and CEO of Vesta said “By integrating Snapdocs’ digital closing infrastructure directly into Vesta, we’re enabling lenders to drive greater efficiency and reduce manual processes.” The integration allows lenders to initiate and manage the full closing process within Vesta, including sending and receiving documents, collecting borrower signatures, handling redraws and tracking real-time status updates. It also supports redraws when loan terms change, helping ensure accuracy without additional steps. Michael Sachdev, CEO of Snapdocs said “By integrating Vesta with our digital closing infrastructure — including advanced eClose capabilities, CD Balancing, automated Funding and Post-Close Quality Control, and an eVault — we eliminate that friction.
GoDocs launches SBA Loan Docs automation eliminating 70% processing time overhead for 7(a) and 504 loans; while ensuring real-time regulatory updates and built-in compliance logic
GoDocs, announced the availability of GoDocs SBA Loan Docs, a powerful new solution that enables lenders to confidently and efficiently generate SBA and state-law-compliant, attorney-quality, lender-ready loan packages for SBA 7(a) Standard and 504 loans. GoDocs SBA Loan Docs eliminates manual bottlenecks by automating the complex documentation process, preserving guaranty eligibility and enabling lenders to scale SBA lending profitably. Whether lenders are expanding into SBA lending for the first time or optimizing an existing program, GoDocs SBA Loan Docs delivers a modern, scalable solution built for today’s compliance landscape. By combining legal-grade precision with intuitive automation, GoDocs empowers financial institutions to close more loans, serve more small businesses, and stay confidently ahead of regulatory change, all while accelerating speed to market. GoDocs SBA Loan Docs is purpose-built for the 7(a) and 504 loan program, offering: Full SOP and state-law compliance, protecting the SBA guaranty; Dynamic document generation, delivering complete packages in minutes, not days; Attorney-drafted documents, continuously updated to reflect policy changes; Built-in compliance logic, reducing audit risk and rework; and Scalable automation, enabling high-volume processing without increasing headcount
Embedded finance platform Parafin launches AI-powered Pay Over Time delivering 10-second credit decisions through LLM banking transactions analysis at point of need – whether at checkout, on invoices, or during payroll runs
Parafin, has launched Pay Over Time, the latest addition to its suite of financing products. Pay Over Time introduces a new payment method that lets small businesses defer payments on essential expenses such as inventory, equipment, payroll, advertising, and rent—directly within the platforms they already use. The product is already live with Gusto as a payroll line of credit. With Pay Over Time, Parafin is broadening access by embedding real-time credit at the point of need—whether at checkout, on invoices, or during payroll runs—helping businesses better manage everyday expenses alongside long-term investments. Unlike conventional lenders who rely on personal credit scores, Parafin uses LLM-powered AI models to analyze raw banking activity in real time, turning thousands of unstructured transactions into a clear picture of business health. This enables credit decisions in as little as 10 seconds with high approval rates, while platforms benefit from seamless integration as Parafin manages underwriting, compliance, and servicing. Vineet Goel, Chief Product and Technology Officer and Co-founder of Parafin. “Pay Over Time empowers platforms to support their customers at the exact moment of purchase, payment, or payroll. With this launch, we’re just getting started on building a comprehensive suite of financial solutions designed for speed, scale, and the realities of small business financing.” In addition to payroll, Pay Over Time is already expanding into use cases such as: Freight booking, Restaurant equipment financing, Dental supply procurement, and Digital advertising
loanDepot seeks dismissal of Maryland class-action lawsuit alleging loan officer compensation violations, claiming plaintiffs with 2.5-3.5% rates lack “concrete injury” standing
loanDepot has filed a motion to dismiss a class-action lawsuit in Maryland alleging violations of loan officer (LO) compensation rules and borrower steering, arguing the plaintiffs failed to show they suffered a “concrete injury.” The lawsuit, filed in July by five borrowers who obtained mortgages from 2019 to 2021, claims loanDepot engaged in a “sophisticated, years-long scheme” to falsify internal documents and federal disclosures to maximize profits ahead of its 2021 initial public offering (IPO). The plaintiffs are suing under the Truth in Lending Act (TILA) and also allege wire fraud, securities fraud and conspiracy. In its Sept. 12 filing, loanDepot argued the borrowers “lack standing” because their loans were not directly affected by the alleged scheme. The plaintiffs, deemed as “far from victims,” received loans with historically low interest rates ranging from 2.5% to 3.5%, the company claimed. A spokesperson for the plaintiffs has not responded to HousingWire’s request for comment. According to the complaint, loanDepot required LOs who couldn’t push higher-cost loans to “transfer” the borrower to an internal loan consultant (ILC) under the false pretense that it was done at the “customer’s request.” But the transfer was described as “fiction,” since the original LO supposedly continued performing the same duties. The firm allegedly punished LOs with reduced commissions if they failed to close loans at inflated rates, or eliminated compensation entirely if they didn’t falsify documentation to conceal the activity. Meanwhile, borrowers were routinely steered into more expensive loans by LOs who were under pressure to offer the highest pricing and faced financial penalties for failing to do so, the lawsuit claims. None of the plaintiffs’ loans were transferred to these ILCs, so they claim they paid higher rates and fees. “Even assuming these assertions are true — and they are not — this alleged ‘scheme’ was not used for Plaintiffs’ loans; it was purportedly used for loans issued to other consumers who ultimately received lower rates based on the alleged TILA violation,” the company said. “Put differently, Plaintiffs’ sole claim in this case rests on the stunning proposition that loanDepot should be held liable under TILA, and the LO Comp Rule specifically, because unidentified loan officers gave unspecified lower interest rates to unidentified borrowers who are neither parties in this case nor members of the proposed class. Neither logic nor law supports that extraordinary theory,” it added. loanDepot also pointed to a three-year statute of limitations for TILA claims. The company added that the plaintiffs provided “scant detail” on the particulars of the alleged scheme by failing to identify a single loan officer or manager. There’s also no information on how “Plaintiffs supposedly ‘discovered’ loanDepot’s alleged fraud,” the firm said. When plaintiffs seek to harm a company’s reputation by asserting sweeping claims of fraud without facts or evidence — and without any explanation for how they know about the supposed fraud — they should not get a second chance,” loanDepot argued.
Channel99 and LinkedIn transform B2B attribution using AI-powered decision engine exposing 80% of previously untraceable website activity through view-through measurement
Channel99, an AI-powered decision engine for B2B marketers, has joined the LinkedIn Marketing Partner Program for B2B Attribution & Analytics, introducing advanced attribution technology for LinkedIn’s paid and organic channels. This integration allows for the correct allocation of “clickless” engagement—previously lost in the “Direct” channel of web analytics—enabling better measurement of campaign ROI tied to pipeline influence. Early results indicate that user engagement attributed to social has increased by 2-4 times with LinkedIn’s Company Intelligence API integration. Existing customers can connect via Single Sign-On (SSO) and quickly start gaining comprehensive attribution insights. CMO of Inovalon, Nick Panayi, noted that this new visibility into social marketing ROI has led to a planned budget shift towards LinkedIn. Channel99 enhances B2B marketing attribution by incorporating view-through attribution, which tracks users who see ads or content before visiting the advertiser’s website. This feature supports a more reliable assessment of campaign effectiveness and allows for easier comparison of pipeline impact across various channels. Chris Golec, Founder & CEO of Channel99, highlighted that acknowledging view-through attribution can potentially enhance ROI metrics by up to 400%. He cautioned that neglecting this aspect would lead marketers to make poorly informed investment decisions, jeopardizing their pipeline targets. Companies can create a free account at www.channel99.com to connect multiple media channels, CRM systems, and intent providers in minutes. The platform also offers AI-driven dashboards for generating recommendations on campaign adjustments and facilitates data integration into customer data platforms (CDPs) and data warehouses.
Mizuho Americas becomes 14th major bank subscribing to Versana’s API-first platform for $6 trillion syndicated loan market straight-through processing optimization
Versana announced that Mizuho Americas, part of Mizuho Financial Group, has joined its centralized, real-time digital data platform as a subscriber. Mizuho Americas (“Mizuho”) has become the 14th top-tier financial institution to join Versana and embrace the modernization of the $6 trillion broadly syndicated loan (BSL) asset class. By leveraging Versana’s API-first platform, Mizuho will optimize key aspects of its loan operations to achieve straight-through processing, reduce costs and improve risk-weighted balance sheet management. Mizuho’s adoption of Versana reflects the bank’s ongoing investment in its digital transformation strategy, implementing cutting-edge technologies to deliver best-in-class services for its clients and shareholders. Mizuho is also adopting the Versana Reconciliation Module into its workflows, which electronically matches lenders’ positions to agents’ golden-source position data. This industry-leading solution reduces discrepancies and improves the accuracy of loan positions used in internal accounting, portfolio management and risk management systems across the loan lifecycle. Matthew Wilson, Managing Director at Mizuho said “Versana’s innovative platform enables us to gain transparency and use better quality data to achieve straight-through processing. By materially streamlining operations and strengthening data integrity, we can better serve our global stakeholders with the most timely and highest quality loan data available.”
DailyPay expands beyond on-demand pay with workplace messaging platform for deskless and hourly workers, targeting 78% absenteeism reduction through enhanced employee engagement and productivity
On-demand pay provider DailyPay has introduced ‘Frontline Communications,’ a new feature integrated into its app aimed at enhancing workplace engagement for hourly employees. Scheduled for an autumn release, this tool addresses communication challenges typically faced by employers of deskless and hourly workers, as traditional email is often ineffective due to varying shift patterns. By integrating communication tools directly into its app, which employees frequently use to check their earned wages, DailyPay seeks to enhance internal messaging. The company has highlighted its position as a high-engagement platform, referencing a commissioned study that ranked DailyPay as the top-adopted financial wellness benefit, with 55% of surveyed employers listing it among their top three most engaged benefits. A key goal of the new solution is to increase employee engagement and reduce absenteeism, with research from Gallup indicating that well-engaged business units experience 78% less absenteeism. Carly Brush, DailyPay’s SVP and general manager for on-demand pay and HCM, emphasized that the expanded capabilities would strengthen the connection between employers and employees, providing real-time access to earned pay and enhancing communication, which in turn boosts engagement and productivity. The new feature complements DailyPay’s existing financial wellness tools, including cash back offers, a credit health product, and savings jars, as the company continues to support hourly workers.
Friendship apps use diversified subscription monetization infrastructure to transform social connections into recurring revenue streams; generated $16 million consumer spending in 2025 with 4.3 million downloads to address loneliness crisis in urban areas
Thanks to online dating apps, the stigma associated with finding connections online has largely faded away. This has welcomed a new wave of apps focused on fostering friendships and building local communities. 1) The app 222 is an iOS-only social events platform that facilitates in-person meetups by pairing a group of strangers based on their personality test results. The app sends invitations to nearby public social events, such as wine bars and comedy clubs. There is then a vetting process, and selected participants are notified on the day of the event. A bonus for those who feel socially anxious: You’re allowed to bring a plus-one. 2) Dating giant Bumble has recently undergone a significant redesign, with an increased emphasis on facilitating group meetups, catering to users’ desires to grow their social circles. Bumble BFF is a feature within the Bumble dating app that allows users to make platonic friends, functioning as a separate mode from the original dating and networking options. 3) Clyx is another emerging app in the group-based social networking landscape, with a strong focus on discovering local events. The social platform helps users find community events by integrating data from platforms like Ticketmaster and TikTok. Additionally, Clyx allows users to upload their contact lists, helping them see which events their friends plan to attend. The app also includes a feature that recommends other users to connect with at these events. 4) Les Amís is a friendship app tailored for women, transgender, and LGBTQ+ individuals that leverages AI to match users based on similar interests and encourages participation in local events, such as pottery classes, book clubs, and wine tastings. Matches are made every Monday, allowing users to chat and plan meetups later in the week. 5) Meetup is helping millions of users connect with others who share similar interests. Users can RSVP to events; join groups that cater to a variety of hobbies, professions, or social causes; and create their own groups and events. They can also chat with group members and post updates and photos from gatherings. 6) European community-building app Meet5 recently launched in the U.S., targeting users over 40 who want to meet new people in their area and participate in group activities such as picnics, concerts, and hiking. Notably, in a short time, there have been around 777,000 U.S. downloads so far. 7) Pie is another one of the newer social apps on the market. It features an AI-driven quiz designed to predict which users are most likely to be compatible with each other. Each person who RSVPs to a Pie event takes a brief personality assessment, and the algorithm organizes attendees into groups of six, who are then added to a group chat within the Pie app, allowing them to interact with each other before the event. 8) Timeleft is a relatively new platform that helps you organize weekly dinner dates with groups of strangers. The app uses a special algorithm to match you with others, taking into account your age, gender, and personality. Users are matched with four other people, but they only learn minor details about them the night before, including their occupations and zodiac signs. 9) Wyzr Friends is an activity-based friendship app designed for adults 40 and older, catering to empty nesters, those who are divorced, and other users seeking to connect with like-minded individuals.
