Answers, first.

A KYC (Know Your Customer) API lets a business verify a customer's identity programmatically — in seconds, during signup — instead of relying on a person reviewing scanned documents. It typically checks a submitted ID or biometric against government, telco, or institutional records and returns a match or no-match decision the calling application can act on immediately. Smile's identity verification product is built as this kind of real-time KYC API: it checks applicants against government, telco, and institutional data sources and forms the foundation of the Verify stage in Smile's broader risk profile.

Business verification, also known as KYB (Know Your Business), is the process of verifying a company's legal registration, ownership structure, and standing before onboarding it as a customer, merchant, or partner. It typically involves checking business registration records, confirming beneficial ownership, and screening the company and its principals against watchlists — the same diligence KYC applies to individuals, extended to legal entities. Smile automates KYB as part of its Verify stage, pulling registration and ownership data through the same API and risk profile used for individual identity verification, so businesses and their owners share one continuous risk record.

KYC (Know Your Customer) verifies individual identities; business verification, or KYB (Know Your Business), checks companies — their registration, beneficial ownership, and standing. Put simply, KYC answers "is this person who they say they are," while KYB answers "is this business real, legally registered, and owned by who it claims." A lender onboarding an individual borrower needs KYC; a lender onboarding an SME, or a platform onboarding a merchant, needs KYB — often alongside KYC on the business's beneficial owners. Smile applies the same Verify-Assess-Monitor framework to both, so an individual's and a business's risk profiles live on shared infrastructure rather than separate systems.

Device fingerprinting builds a unique signature for the phone or browser behind an application, using signals like hardware identifiers, OS configuration, and behavioral patterns — so a fraud team can tell whether an applicant is using a genuine, unique device or an emulator, a spoofed device, or one device reused across many applications. That last pattern — one device linked to dozens of 'different' applicants — is a strong signal of coordinated fraud rings or bot-driven signups. Smile also layers in behavioral biometrics (typing and touch patterns) and remote-access-tool detection, which catches a different fraud pattern entirely: a fraudster using tools like TeamViewer or AnyDesk to control a victim's device during a live session, a common tactic in account-takeover and social-engineering scams. Combined with phone and email digital footprint signals in its Assess stage, device-level fraud is caught before an application reaches underwriting, not after.

Credit decisioning is faster and more accurate when a lender has verified identity, income, and risk signals in one place instead of stitching together data from multiple vendors. Smile supplies real-time identity, income, and employment data in its Verify stage — including alternative data sources like payroll, HR, and gig-platform records for thin-file applicants who lack traditional credit history — then feeds that same verified profile into fraud and risk scoring in its Assess stage. The result: lenders get an approve, decline, or review recommendation without collecting or reconciling the same applicant data twice across separate systems.

AML (anti-money laundering) and watchlist screening checks an applicant or business against government sanctions lists, politically exposed persons (PEP) databases, and blacklists to catch individuals or entities that are legally prohibited from being onboarded, or that carry elevated regulatory risk. Screening once at onboarding isn't enough on its own, because a customer can appear on a watchlist — or change account details in a way that signals takeover — after they've already been approved. Smile runs initial AML and watchlist screening in its Assess stage, at onboarding and on every account or beneficiary change, then re-screens on an ongoing basis in its Monitor stage, so a new match surfaces as an alert later rather than going undetected.

A fraud and risk score is a single output — typically approve, decline, or review — that combines many underlying signals, such as identity verification results, device and location intelligence, income data, and behavioral patterns, into one decision a business can act on without manually weighing each signal itself. Smile's fraud and risk scoring product sits in its Assess stage and draws on 700+ signals across identity, device, location, behavior, and income, designed to run alongside a lender's existing credit or risk model rather than replace it — recalibrating decisions on both sides, catching risk an existing model misses and surfacing good applicants it would otherwise turn away.

Early payment default (EPD) is when a borrower misses a payment shortly after a loan is disbursed, often within the first one to three installments — a strong early indicator of both fraud and poor underwriting, since it surfaces risk that credit history alone didn't catch. Digital identity and behavioral signals correlate with EPD risk in ways a traditional bureau score can't see: how long an applicant's digital identity has existed, how far their device's location sits from their stated address, and how much account activity they've shown historically. Smile's fraud and risk scoring product, part of its Assess stage, surfaces these signals as EPD risk indicators, so a lender can factor them into a decision before disbursement rather than discovering the risk after the fact.

Loan stacking is when a borrower applies for and takes out multiple loans from different lenders within a short window — often in one sitting — usually without disclosing the other applications, leaving each lender unaware of the borrower's total exposure. It's a leading cause of non-performing loans in fast-approval lending, because a borrower's debt-to-income ratio at underwriting no longer reflects reality by the time all loans fund. Smile detects loan stacking in its Monitor stage by tracking application velocity across its network and flagging blacklist or repeat-applicant matches in real time, before a second or third loan is approved.

Banking regulators in many markets are introducing anti-fraud legislation that mandates specific detection categories — transaction velocity monitoring, account-takeover and change-of-details monitoring, geolocation checks, blacklist screening, and behavioral anomaly detection. Smile's Assess stage covers the real-time device, geolocation, and behavioral-anomaly signals these regulations typically require, while its Monitor stage covers ongoing transaction-velocity and blacklist re-screening — through the same API used for identity, income, and AML checks, adapting to whichever regulatory framework applies in your market.

Smile connects to a lender's or fintech's existing underwriting flow through a single REST API spanning its Verify, Assess, and Monitor stages — identity and income verification, device intelligence and fraud scoring, and AML/watchlist and loan-stacking checks. Rather than integrating five or more point-solution vendors — each with its own contract, API, and data format — a team calls one API and receives one consistent risk profile per applicant, which plugs into an existing decision engine or underwriting workflow with a single integration.

Yes. Smile is built to comply with the data privacy regulations applicable in each market it operates, adapting consent and data-handling practices — including data portability rights where required by local law — to that market's requirements, backed by consent-based data collection across every stage of Smile's Verify, Assess, and Monitor framework.