Credibly AI Underwriting Patent

Credibly has obtained a U.S. patent for its AI-driven SMB lending platform, enhancing automated underwriting and capital origination using generative AI, neural networks, and machine learning.

The patented system automates and improves underwriting and funding decisions, transforming the manual process into a faster, more accurate, and scalable operation.

Key details regarding the patented technology:

• The platform features a proprietary search engine that analyzes incoming applications using historical data, generative AI, and machine learning, driving improved decision-making speed and underwriting precision.

• It automates risk assessment, leveraging AI-trained models to classify applications into credit bands, matching offers to applicant profiles more precisely and efficiently.

• This automation reduces the workload for human underwriters, mitigates manual errors, increases funding speed, and supports better-matched financing offers to SMBs needing quick capital access.

This accomplishment further establishes Credibly as a leading GenAI-powered fintech lender, with a track record of more than $2.6 billion in capital provided to over 50,000 SMBs since 2010.

Has Credibly Secured an Undeniable Competitive Moat with its Generative AI Underwriting Patent?

Management sees this patented system as a unique search engine using GenAI, Multiple Neural Networks, and Machine Learning for large-scale working capital origination.

Securing AI technology patents is vital for market protection and enhancing investment value.

This process uses a proprietary search engine with trained neural networks and historical data to streamline loan application reviews.

What Immediate Economic Impact Do the Patented AI Efficiencies Deliver?

The most compelling proof of the innovation's utility are highly disruptive: Credibly reports achieving an 85% decrease in operational costs through automation and processing 4x the volume of submissions without increasing review time. These figures translate directly into superior speed and profitability that traditional models simply cannot match.

• This new automation streamlines the entire application review process, replacing the traditional requirement for individual underwriters to manually evaluate and verify every single detail of an application.

• By integrating this patented GenAI technology, Credibly is now capable of assessing borrower affordability with greater precision, strengthening risk management protocols, and generating financing offers that are meticulously matched to the SMB's specific risk profile.

• Beyond mere speed, AI's objective, data-driven approach enhances core business outcomes, leading to increased underwriting accuracy and faster funding decisions, delivering a seamless experience for SMBs seeking prompt capital access.

• The adoption of AI and machine learning in financial credit decisions generally allows lenders to make smarter, more comprehensive underwriting decisions by analyzing thousands of data points and incorporating non-traditional metrics, a necessity in the modern lending landscape.

Does Credibly’s Decade-Plus Scale and History Validate its Risk Management Acumen?

Founded in 2010, Credibly has proven its resilience by providing over $3 billion in capital to more than 60,000 SMBs.

• The company has consistently prioritized risk management and a culture of compliance throughout its operations since launching, a strategy that allowed it to achieve sustained expansion and record exceptional origination numbers.

• Its operational longevity includes successfully navigating the challenging economic environment of the COVID-19 pandemic, demonstrating robustness during periods of significant market stress.

• As a testament to its disciplined approach to portfolio management, Credibly made history in 2017 by becoming the first company in its sector to acquire the servicing rights to another alternative lender’s $250 million portfolio.

• Credibly provides a diverse offering of funding options, including balance sheet, syndication, and off-balance sheet financing solutions, catering to a wide array of business capital needs.

Does Credibly’s 85% Cost Reduction Claim Lack Essential Portfolio Performance Validation?

Credibly's claims of an 85% reduction in operational costs and 4x faster processing due to automation are meaningless without proven credit performance metrics. In high-risk SMB lending, speed and cost savings are detrimental if they increase default rates or loss severity.

• The Missing Link: The core announcement documents confirming the issuance of the patent for the proprietary computing system using Generative AI (GenAI) and Multi-Class Neural Networks focus on the internal operational gains of the technology—namely increasing underwriting accuracy and streamlining the typically manual, time-intensive evaluation process. Critically, there are no public information that provide quantifiable data isolating the net charge-off (NCO) rates, default rates, or loss severity metrics achieved directly by this patented GenAI platform when compared against Credibly’s previous, non-AI-driven underwriting methods.

• The Underwriting Mandate: The primary purpose of integrating AI and Machine Learning (ML) into underwriting is fundamentally to perform smarter, faster credit assessment by moving beyond obsolete criteria like sole reliance on traditional credit scores. The AI is designed to integrate proprietary databases, leverage trained neural networks, and assess factors beyond credit scores, including cash flow trends, payment histories, and macroeconomic indicators, which should theoretically lead to tailored loan offers and strengthened risk management.

• Validation is Non-Negotiable: For institutional executives evaluating the stability of the platform as a debt or equity investment, the claimed operational cost savings must be unequivocally proven to improve, or at least maintain, portfolio health. The integration of patented GenAI technology is explicitly aimed at ensuring affordability assessment is more precise and offers are better matched to the borrower's risk profile. Without transparent reporting on default rates and loss metrics, the efficacy of the patented risk component remains purely theoretical.

How Does Credibly’s AI Performance Benchmark Against Proven Industry Competitors like Upstart?

In the absence of isolated AI performance data directly from Credibly concerning its patented GenAI platform, market analysts must rely on the established transparency and performance metrics reported by competitors who have successfully linked AI adoption to superior credit risk management.

• Upstart's Quantifiable Risk Moat: Upstart, a leading AI lending platform, provides a clear benchmark by linking its automation efficiency to significantly enhanced risk separation and expanded borrower access. Upstart’s AI models achieved 6x greater risk separation between its highest and lowest risk grades compared to the 2x differentiation provided by traditional FICO models. This level of stratification is the gold standard for validating AI efficacy.

• Approval vs. Risk Trade-Off: Leading AI platforms demonstrate that speed does not necessitate sacrificing credit quality. Upstart reports that its models enable partner lenders to approve 101% more borrowers at significantly lower Annual Percentage Rates (APRs)—38% lower—than traditional models, all while successfully maintaining portfolio health. Furthermore, Upstart achieved 89% accuracy in predicting 90-day defaults, vastly outperforming traditional FICO models (72% accuracy).

• Operational Integration with Risk: Upstart’s operational enhancements, such as achieving a 91% loan automation rate, are continuously tied back to risk management improvements, including a 13% year-over-year reduction in roll rates from delinquency to charge-off. This demonstrates the transparent competitive standard required to prove that automation reduces cost without escalating credit risk.

What Do the Available Aggregate Portfolio Performance Metrics Indicate About Credit Quality?

While Credibly’s specific GenAI platform performance remains non-disclosed in the sources, the broader institutional lending platform, Propel Holdings (which employs AI/ML underwriting and operates several SMB lending brands like CreditFresh and MoneyKey, alongside Credibly's historical funding records), provides aggregate metrics that offer partial insight into the portfolio's overall credit discipline.

• Net Charge-Off Rate Context: The aggregated data shows that the Net Charge-Offs (NCOs) as a percentage of Average Combined Loan and Advance Balances (CLAB) actually increased to 12% for the three and six months ended June 30, 2025, up from 11% in the corresponding 2024 periods. For a rapidly expanding portfolio—Total Originations Funded increased by 35% year-over-year—a rising NCO ratio warrants heightened scrutiny, especially when simultaneously claiming breakthrough technological efficiency.

  • The company states that the 12% NCO rate is considered "well within our target range" and reflective of strong credit performance amid significant growth.

  • This ratio is projected to decrease over time as AI capabilities enhance underwriting and facilitate products for lower credit risk consumers through variable pricing and graduation programs.

  • The platform’s lending brands are continuously refining underwriting strategies and updating AI models to manage risk, ensuring that growth is profitable even when operating under continued macroeconomic uncertainty.

• Provision for Loan Losses Relative to Revenue: The overall Provision for Loan Losses (PFL) and other liabilities—which includes forward-looking expected credit losses (ECL) under IFRS 9—was maintained at 50% of revenue for Q2 2025, consistent with Q2 2024.

  • The overall stability of the PFL-to-revenue ratio, despite massive growth, is internally cited as indicative of strong unit economics and strong credit quality.

  • In theory, the variable pricing functionality enabled by AI is expected to compensate for any marginal decline in Annualized Revenue Yield by generating a relative decline in PFL going forward.

• Need for Isolated AI Validation: The company has been strategically emphasizing technology, including launching the Credibly.ai innovation platform to showcase Generative AI use cases. While these operational advancements suggest capability, the lack of transparent, isolated credit performance data for the patented GenAI models prevents the definitive verification that the 85% cost reduction is sustained by superior risk management, as opposed to simply accelerating the conversion funnel. Analysts require direct validation that the core function of the proprietary AI—to generate better-matched financing offers—is definitively improving the bottom line beyond mere operational savings.

1-Minute Video: Why SOS Data Mismatch is the #1 Reason Legit Businesses Get Denied?

SOS Data Mismatch is the #1 Reason for Application Denials?

The "vast majority of denials" stem from incongruent data between loan applications and Secretary of State records. This isn't about actual fraud—it's about legitimate business evolution that never gets reflected in state filings.

Why This Matters:

• Your automated fraud detection is optimized for speed and risk avoidance, not accuracy

• Every false positive denial represents lost origination revenue and wasted acquisition costs

• Your marketing spend is being burned on leads your systems will systematically reject

• Portfolio growth targets are being artificially constrained by preventable data friction

Here's the Lending Reality:

Small business owners establish their LLC, then sequentially acquire:

1. Business address (often different from formation address)

2. Business phone number (post-formation)

3. EIN and bank accounts (after entity creation)

4. Operating licenses and permits

None of these updates trigger SOS filing updates. The entrepreneur doesn't think about it, and state filing requirements don't mandate it for most changes.

The solution isn't to lower standards or ignore fraud risks. The solution is to improve verification accuracy so underwriters can focus on credit risk, not data hygiene issues.

Your competitive advantage depends on:

• Converting qualified borrowers your competitors are denying

• Funding deals faster with lower operational costs

• Maintaining portfolio quality and fraud loss rates

• Scaling origination without proportional headcount increases

The question isn't whether to invest in automated verification. The question is: How much revenue are you willing to lose while your competitors figure this out first?

Schedule a FREE demo call

Headlines You Don’t Want to Miss

mPWR Accelerates Latin American Expansion with Strategic Acquisition of Mexican Fintech Kredeo

mPWR, a mobile-first fintech company specializing in digital lifestyle services, announced its acquisition of Mexican fintech Kredeo through an all-equity deal granting it 100% ownership. The move provides mPWR immediate access to Mexico’s regulatory licenses, lending infrastructure, and technology, marking a major step in its strategy to expand across Latin America and empower underserved consumers with accessible financial products.​

Kard secures $15m growth capital from Trinity Capital

Trinity Capital has committed $15 million in growth capital to fintech company Kard to accelerate the expansion of its commerce media network and personalized rewards platform. The funding will enable Kard—already processing over $10 billion in monthly transactions and reaching tens of millions of consumers—to scale operations, enter new markets, and further develop its AI-driven, merchant-funded rewards ecosystem.

Carvana plunges as investors respond to another subprime lender's bankruptcy filing

Carvana shares plunged over 8% as investors reacted to the bankruptcy filing of PrimaLend, a subprime auto lender that finances dealers serving borrowers with low credit scores, raising fears of broader contagion in risky auto credit markets. Although Carvana has no direct connection to PrimaLend, the collapse renewed scrutiny of its heavy exposure to near-prime and subprime loans, with analysts warning that rising defaults could threaten its growth-dependent business model.