Smile API Weblog

Smile API Named a Finalist in the Global FinTech Hackcelerator 2026Smile API is excited to share that we have been selected as a finalist in the Global FinTech Hackcelerator 2026, organised by the Global Finance & Technology Network (GFTN) in partnership with the Monetary Authority of Singapore (MAS).As part of GXS Bank’s Adaptive Risk Modelling Systems challenge, we are exploring how lenders can bring credit and fraud information together to build a more complete picture of the customers they serve.We look forward to showcasing our approach at the Singapore FinTech Festival on 18 November 2026.Connecting credit and fraud assessmentEvery lending application raises two connected questions: Can this person repay, and is the application genuine?To answer them, lenders may consider different types of information. Income and employment data can help inform affordability assessments, while identity and fraud indicators can help teams assess whether an application warrants closer review.Bringing these perspectives together could give lenders a clearer basis for making decisions—particularly when a customer’s circumstances are not fully reflected in a traditional credit record.Risk can also change over time. A customer’s income or employment situation may shift after approval, making it valuable for lenders to consider how relevant information can inform decisions throughout the customer relationship.Smile’s proposed approachIn response to the Adaptive Risk Modelling Systems challenge, Smile is exploring how relevant identity, income, employment and fraud signals could be brought together into a customer risk profile that can be updated as new information becomes available.For example, user-permissioned income and employment information could help a lender better understand a gig worker whose earnings come from multiple platforms. Considered alongside other relevant risk indicators, this information could support a more informed assessment.Through the Hackcelerator, we aim to refine and validate our proposed approach against real lending needs. Our goal is to help lenders better understand risk while making more informed decisions about the customers they serve.Looking ahead to Demo Day - As finalists, we look forward to learning from industry mentors, engaging with GXS Bank and refining our approach alongside other teams working to solve challenges across financial services.We’re proud to be part of the Global FinTech Hackcelerator 2026 and excited to share our progress at the Singapore FinTech Festival.Meet Smile at the Singapore FinTech FestivalSmile will be at the Singapore FinTech Festival from 18 to 20 November 2026, with our Hackcelerator presentation on 18 November.If you’re attending and would like to learn more about our approach to connected credit and fraud assessment, we’d love to connect.Follow Smile API on LinkedIn for updates as we prepare for Demo Day, or visit www.getsmileapi.com to learn more about Smile.

How an Income Estimator API Adds Context to Thin-File Lending Decisions Income is an important part of lending decisions, but it is not always easy to assess early in a digital application. Some applicants may not have a traditional payslip on hand. Others earn through freelance work, online platforms, or several sources. Asking every applicant for more documents can slow onboarding and create unnecessary friction. An income estimator API gives lenders another way to assess an application before deciding whether more verification is needed. Smile API’s Income Estimator uses behavioral signals to estimate the likelihood that a user belongs to a higher-income segment. With the applicant’s consent, a lender submits an email address and country code. The API returns a synchronous response that can be added to an existing onboarding, credit, or risk workflow. The result is not an exact salary calculation. It is an additional signal designed to help lenders segment applicants, prioritize reviews, and decide when verified income documents are still necessary. Why traditional income checks can create friction Traditional income verification remains important, especially when a lender needs definitive evidence. But requiring documents too early or from every applicant can introduce several challenges: Applicants may need to find, download, or photograph a payslip or bank statement.Submitted documents may require manual review.Non-salaried or gig-economy workers may not have standard proof of income.Extra steps can increase the time required to complete an application.Thin-file applicants may have limited conventional data even when they have the capacity to repay. The goal of an income estimation signal is not to remove verification. It is to help a lender use verification more selectively. When the estimator returns a weaker or less certain result, the lender may request more evidence. When the signal is stronger, the lender can use it alongside its other data and policies to determine the appropriate next step. What is Smile API’s Income Estimator? Smile API’s Income Estimator is an email-based income estimation API for lenders and fintech companies. It predicts the likelihood that a user belongs to a higher-income group and maps that prediction to a predefined income tier. The API returns four main outputs: 1. Probability Probability is the model’s primary predictive output. It represents the likelihood that the user belongs to the higher-income segment. For decisioning, this value carries more information than the simplified tier. Two applicants may fall into the same tier but receive different probabilities, allowing a lender to distinguish between a stronger and a weaker model signal. 2. Income tier The tier translates the model output into a simpler category, such as Low, Middle, or High. It can support customer segmentation, user-interface displays, portfolio reporting, or straightforward business rules. A tier should not be interpreted as the applicant’s verified or exact monthly income. 3. Income range The income range shows the predefined reference boundaries associated with the returned tier. It provides context for the category, but it is not a personalized salary estimate. Applicants assigned to the same tier receive the same reference range. 4. Confidence Confidence describes the model’s certainty relative to its internal decision boundary. It do

Credit Capacity: Rethinking How Lenders Set Loan Limits For years, lenders have relied heavily on credit scores to answer a fundamental question: Should we lend to this borrower? But approval is only part of the decision. Credit capacity asks another equally important question: How much can this borrower realistically afford to repay? A borrower may appear creditworthy based on past repayment behavior, yet that does not automatically mean every loan amount is suitable. By combining traditional credit information with verified income and employment data, lenders can build a clearer picture of a borrower's present financial capacity—and use that insight to make more informed lending decisions. Creditworthiness and Credit Capacity Are Not the Same A traditional credit score is primarily designed to estimate credit risk based on available historical information. It can help lenders understand whether someone has demonstrated responsible repayment behavior. But historical behavior does not always provide a complete picture of the borrower's situation today. Consider two borrowers with similar credit profiles. One earns a stable monthly salary and has remained with the same employer for several years. The other has recently experienced a significant drop in income. Their credit histories may look similar, but their ability to comfortably manage the same new monthly payment could be very different. This is where credit capacity adds another layer to underwriting. Instead of relying on a single risk indicator, lenders can consider factors such as: Current and verified income Employment status and history Income consistency Existing financial obligations Expected repayment amount relative to available income The goal is not to replace credit scores. It is to supplement them with information that helps lenders determine an appropriate level of credit. The World Bank notes that alternative data—including information about income, transactions, sales, taxes, and bill payments—can provide useful indicators of a borrower's ability or willingness to repay. Why Loan Amount Matters as Much as Loan Approval A binary approve-or-decline model can overlook an important middle ground. Imagine a borrower who may comfortably repay a PHP 20,000 loan but could struggle with a PHP 75,000 obligation. Simply identifying that borrower as “low risk” does not necessarily determine which loan amount is appropriate. Credit capacity can help lenders think beyond eligibility and toward affordability. This may support more granular decisions around: Loan limits Credit-line assignments Repayment periods Installment amounts Product eligibility For lenders, better alignment between loan size and repayment capacity can support more responsible portfolio management. For borrowers, it can help reduce the risk of receiving more credit than their current finances can reasonably support. This approach is especially relevant in markets where many consumers have limited traditional credit histories. The World Bank has highlighted how alternative data can complement credit bureau information and provide lenders with a better understanding of thin-file borrowers.  Verified Income Data Makes Capacity More Visible The challenge is that credit capacity is difficult to estimate when lenders cannot reliably see a borrower's current income and employment situation. Manual processe

Loan Servicing Philippines: What Happens After ApprovalLoan servicing in the Philippines has a blind spot — and it begins the moment a borrower signs their approval letter. For most financial institutions, the verification process ends at disbursement. The income was checked, the employment was confirmed, the credit decision was made, and the money moved. What happens next, from a data standpoint, is largely silence.That silence is where default risk lives.The reality of lending is that a borrower's financial situation on the day they are approved is not a permanent condition. Employment ends. Income drops. A second job disappears. A company downsizes. Life changes — and when it does, lenders are almost always the last to find out. By the time a missed payment appears on a statement, the early intervention window has already closed, the borrower is already under financial stress, and the cost of resolution — restructuring, collections, write-offs — has already begun to accumulate.Real-time employment and income data is changing this dynamic. And the lenders who understand its value beyond the application stage are building a significant and largely invisible competitive advantage in portfolio risk management.The Loan Lifecycle Nobody Talks AboutThe Philippine lending content landscape — and frankly, the global one — is dominated by front-end thinking. Acquisition, verification, decisioning, disbursement. These stages receive the most product investment, the most operational attention, and the most industry conversation.But a loan's true risk profile plays out over its entire term. A 12-month personal loan has 12 months of repayment risk. A 36-month housing loan has 36 months. A microfinance cycle loan that renews quarterly has risk that compounds across every renewal. The verification data gathered at origination tells a lender what a borrower looked like at one specific moment in time — not what they look like six months later when their employer restructures or their contract ends.This is the loan servicing gap: the absence of ongoing, real-time borrower financial intelligence between the point of approval and the point of default.Why Traditional Loan Monitoring Falls ShortMost Philippine lending institutions currently rely on one of two signals to identify at-risk accounts during the servicing phase:Payment behavior — a missed payment or a late payment triggers a flag. This is reactive by definition. By the time the flag appears, the borrower has already defaulted.Periodic contact — loan officers call or message borrowers at set intervals to check in. This is resource-intensive, inconsistent, and entirely dependent on the borrower being forthcoming about changes in their financial situation — which, understandably, many are not.Neither approach gives lenders advance warning. Neither approach allows for intervention before the financial stress has already translated into a missed payment. And neither approach scales efficiently as a portfolio grows.The result is a servicing model that is fundamentally reactive — designed to respond to default rather than prevent it.What Real-Time Employment Data Makes PossibleConsumer-permissioned employment and income data, accessed with the borrower's ongoing consent, transforms loan servicing from a reactive process into a proactive one. Instead of waiting for a missed payment to signal that something has changed, lenders can monitor

Microfinance Verification Philippines: Scale SmarterMicrofinance verification in the Philippines sits at the heart of the country's most critical — and most complex — lending challenge. Serving millions of micro-borrowers and informal workers across the country, microfinance represents the backbone of financial inclusion in the Philippines. It also serves the borrowers who are hardest to assess through traditional means: the self-employed, the informally employed, the underbanked, and those with little to no formal credit history.For microfinance institutions (MFIs), rural banks, cooperatives, and NGO-linked lending programs, the fundamental tension has always been the same: how do you scale responsibly when your borrower base is the one the financial system was least designed to serve?Employment and income data is changing the answer to that question.Understanding the Microfinance BorrowerBefore exploring solutions, it's worth being precise about the challenge. Microfinance borrowers in the Philippines are not a monolithic group, but they share several characteristics that make traditional credit assessment difficult.Many are self-employed — sari-sari store owners, market vendors, small-scale farmers, informal traders. Their income is real and often consistent, but it doesn't arrive in neat monthly deposits from a single employer. It fluctuates with seasons, with foot traffic, with market conditions. A payslip doesn't exist because a payroll system was never involved.Others are contractual or project-based workers whose employment tenure with any single organization is short, making employment verification through traditional channels — phone calls to HR departments, certificate of employment requests — impractical or impossible.And then there's the credit history problem. Many microfinance borrowers are applying for their first formal loan. There is no credit bureau record to consult. There is no history of credit card payments or bank loan repayments to analyze. By traditional credit scoring standards, they are invisible — even if they are among the most financially disciplined people in their communities.The Verification Methods MFIs Have Relied On — And Their LimitsMicrofinance in the Philippines has developed creative approaches to the credit gap problem. Group lending models, where a small cohort of borrowers guarantees each other's loans, have historically been effective at managing default risk through social accountability. Community-based credit officers who know borrowers personally have served as a form of human verification.These methods work — up to a point. The problem is that they don't scale. As MFIs grow their portfolios, as borrower bases expand beyond tight community networks, and as digital lending channels reduce the face-to-face interaction that community-based verification depends on, the old methods become bottlenecks.Manual document review is slow and error-prone. Community credit officers are expensive to deploy at scale. And group lending, while effective for small portfolios, introduces its own operational complexity when the borrower base numbers in the thousands.The result is a ceiling on responsible growth: MFIs that want to expand their impact — and their portfolio — find themselves constrained by verification infrastructure that was designed for a different era. What Employment and Income Data Adds to the Microfinance PictureHere is where

The Borrower Experience Gap in Philippine LendingThe borrower experience gap during a loan application in the Philippines is broken — and it's costing lenders far more than they realize. While financial institutions spend enormous energy competing on interest rates, cashback rewards, and loan limits, a quieter, more damaging battle is being lost somewhere else entirely: inside the application process itself.A borrower who abandons your application isn't just a missed loan. They're a marketing spend that never converted, a relationship that never started, and a customer who will likely tell others about the friction they encountered. And yet, most lenders still treat the application process as an afterthought.It's time to close the gap.What Is the Borrower Experience Gap?The borrower experience gap is the chasm between what a borrower expects when they apply for a loan and what they actually encounter. In 2025, a borrower's expectations have been shaped by the apps they use every day — seamless, fast, intuitive, and requiring minimal repetitive input. What they often encounter in a loan application is the opposite: long forms, repeated data entry, document uploads, waiting periods, and uncertainty about where their application stands.This gap doesn't just frustrate borrowers. It kills conversions. Industry data shows that loan application abandonment happens at every stage of the funnel — but the steepest drop-offs happen exactly where verification friction peaks: income verification and employment checks. On average, only 35–50% of applicants who start a loan application actually complete it. That means lenders are losing up to half of their interested pipeline — not because their rates are wrong, but because their process is.Where the Friction Actually HappensTo fix the borrower experience, it helps to understand exactly where it breaks down.Stage 1: The Form Itself Long, repetitive application forms are the first filter — and not in a good way. When a borrower is asked to manually input employer name, employment start date, monthly income, and contact details for their HR department, every additional field is a small barrier. For mobile-first borrowers — which is the majority in the Philippines — typing out lengthy data on a small screen is enough to prompt a quiet exit.Stage 2: Document Submission This is where abandonment spikes. Being asked to upload clear photos of payslips, a certificate of employment, a valid ID, and sometimes additional supporting documents introduces friction at exactly the moment when a borrower's initial enthusiasm is starting to cool. If even one document is unclear, missing, or flagged, the process stalls — and many borrowers never return.Stage 3: The Waiting Period After submission, borrowers enter the least comfortable part of the experience: not knowing. In a world where food deliveries are tracked in real time and bank transfers clear in seconds, being told "we will get back to you in 3–5 business days" feels jarring. Uncertainty breeds disengagement. And for borrowers who applied out of urgency — which is most of them — a multi-day wait often means they've already found another solution.Stage 4: Re-Submission Requests When a manual reviewer flags an issue — a payslip that's hard to read, an employer phone number that isn't answered, an income figure that doesn't match the declared amount — the borrower receives a request for additional d