Underwriting Research Initiative

Expanding Access to
Responsible Credit

GoaKarma evaluates whether additional financial behavior signals can help lenders identify creditworthy applicants who lack traditional credit history.

We work with lending institutions to conduct observational underwriting analysis without changing their existing approval process.

For Lending Institutions → About the Research
Research Overview
1.4B Adults globally without access to formal credit systems
Observational Analysis only — no lending decisions made by GoaKarma
No changes required To your existing underwriting process
MIT BIG.AI Conference Accepted presenter · April 2026 · MIT Initiative on the Digital Economy
Observational Research Model Shadow analysis only — no interference with live underwriting decisions
Privacy-First Data Collection Consent-based signals · no bureau reporting · no PII retention
Institutional Collaboration Designed for lending institutions · no system integration required
1.4B Global Credit Gap Adults without formal credit history (World Bank Global Findex)
MIT BIG.AI @ MIT Accepted — April 2–3, 2026 · Samberg Conference Center, Cambridge
Shadow Observational Model No interference with live lending decisions
No changes Privacy First Consent-based data. No bureau reporting.
The Industry Problem

The Thin-File Credit Gap

Millions of financially active individuals are declined for credit each year — not due to poor repayment behavior, but due to limited formal credit history. Traditional underwriting tools are designed for borrowers with established bureau records, leaving lenders unable to confidently evaluate otherwise viable applicants.

This gap affects both lenders and borrowers: institutions forgo potential performing loans, while responsible borrowers remain excluded from the formal financial system.

1.4 Billion adults worldwide lack formal credit history  —  World Bank Global Findex

Limited Bureau Coverage

Current underwriting relies heavily on bureau records that don't exist for a significant portion of the population, particularly in emerging markets and among younger borrowers.

Viable Borrowers Declined

Lenders must decline applicants who may be creditworthy — not because of demonstrated poor behavior, but because there is insufficient data to evaluate them under existing frameworks.

Institutional Blind Spots

Standard risk models were built for a different era of financial participation. New behavioral data exists but lacks structured evaluation within formal credit assessment.

Portfolio Opportunity

For lending institutions, this gap represents both risk and opportunity. Identifying performing borrowers within the declined pool could meaningfully improve portfolio outcomes over time.

What GoaKarma Is

An Underwriting Research Initiative

GoaKarma studies whether certain observable financial behaviors — such as payment consistency and digital financial activity — correlate with repayment outcomes over time.

Our objective is not to replace existing underwriting methods, but to determine whether additional information can assist lenders in responsibly expanding approvals while maintaining portfolio quality.

GoaKarma conducts analysis, evaluation, and observational review. We do not make approvals, set rates, or originate credit.

Traditional Credit
Behavioral Model GoaKarma
Data SourceCredit Bureau & Bank Records
Data SourceBehavioral & Community Data
Risk SignalPast Borrowing History
Risk SignalReal-Time Behavior Signals
LogicStatic Credit Scores
LogicAdaptive AI Scoring
AccessRequires Credit History
AccessOpen to All Applicants
Capital FlowBank to Borrower
Capital FlowCommunity to Borrower
FeedbackSlow Bureau Updates
FeedbackContinuous Learning
Concept Overview

How GoaKarma Fits Into the Lending Process

A simplified flow showing where behavioral analysis enters the underwriting pipeline.

01
Borrower
Application
02
Traditional
Underwriting
03
Declined
Applicant Pool
GoaKarma
04
Behavioral
Analysis
05
Institutional
Insights
Standard pipeline flow
GoaKarma entry point
GoaKarma stage
Engagement Model

How the Evaluation Process Works

A structured, four-stage process designed to provide insight without disrupting existing institutional workflows or approval policies.

01

Normal Underwriting

A lender completes its standard underwriting process. No change to existing policies or systems is required at this stage.

02

Application Referral

Applications already declined may be shared with GoaKarma for independent analysis, at the institution's discretion.

03

Observational Review

GoaKarma conducts an offline, anonymized observational review of behavioral and financial signals.

04

Aggregate Findings

Aggregate findings are reported to the institution. No lending decisions are made by GoaKarma at any stage.

Participation does not require any change to a lender's existing approval policies.

Institutional Engagement

For Lending Institutions

GoaKarma does not replace your underwriting process.

GoaKarma provides a structured method to evaluate whether additional data signals may identify repayable borrowers among applicants who would otherwise be declined.

  • ✓No change to your underwriting process
  • ✓No obligation to approve additional applicants
  • ✓No operational integration required initially
  • ✓Observational analysis only
  • ✓Aggregate, anonymized reporting

GoaKarma does not originate loans, set interest rates, or approve borrowers. Lending decisions remain entirely with the institution.

Request an Introductory Discussion →
What We Offer
S
Supplemental AnalysisReview of behavioral signals on referred declined applications.
R
Rejected Applicant ReviewIndependent, offline examination of the declined applicant pool.
P
Optional Pilot EvaluationStructured cohort monitoring and performance tracking over time.
Research Approach

Methodology & Evaluation Approach

GoaKarma analyzes anonymized behavioral indicators and monitors aggregate outcomes over time. Cohort performance is evaluated using observed repayment behavior rather than theoretical modeling alone.

Our approach emphasizes cautious validation and ongoing review. The goal is to understand whether additional signals can meaningfully complement traditional credit assessment — not to displace established underwriting discipline.

Evaluation Framework
◎
Behavioral indicator analysis across financial activity patterns
◎
Cohort observation and longitudinal performance tracking
◎
Outcome measurement against actual repayment data
◎
Continuous model monitoring and ongoing structured review
Validation Approach
◎
Shadow analysis — no interference with live underwriting decisions
◎
Hypothesis-driven evaluation: testing correlation, not asserting causation
◎
Performance tracking with explicit uncertainty acknowledgment
◎
Transparent reporting of cohort findings to participating institutions

GoaKarma seeks to evaluate whether alternative behavioral indicators correlate with repayment outcomes over time. We are testing a hypothesis, not marketing a solution.

Early Research Indicators

Preliminary Signal Analysis

The following charts reflect early-stage observational findings from GoaKarma's research cohorts. These results are preliminary and subject to ongoing validation. They are presented to illustrate the nature of the hypothesis being tested — not to assert proven outcomes.

Discriminative Performance
Predictive Signal Comparison
Receiver Operating Characteristic (ROC) — Cohort A, n=1,240
GoaKarma Model — AUC 0.81
Baseline Model — AUC 0.67
Random Chance
Preliminary data. Ongoing validation in progress. Results may change.
Repayment Outcomes
Observed Repayment Outcomes by Signal Group
12-Month Repayment Rate — Declined Applicant Cohorts
GoaKarma High Signal
GoaKarma Low Signal
Traditional Baseline
Preliminary data. Ongoing validation in progress. Results may change.

Research Caveat: These charts represent early-stage observational findings from a limited research cohort. GoaKarma does not claim that these results are statistically definitive or commercially validated. All figures are subject to revision as the study progresses. This research is presented to illustrate the hypothesis under investigation, not to market a proven model.

Responsible Use

Governance & Responsible Use

Credit decisions carry significant real-world consequences. GoaKarma approaches analysis with a focus on transparency, structured oversight, and measured evaluation. We understand that influencing credit carries institutional responsibility.

Data Privacy

All analysis is conducted on anonymized data. Personal information is never retained beyond observational review or shared across institutions.

Auditability

Our evaluation process is fully documented and available for institutional review, including methodology, cohort composition, and outcome tracking.

Fairness Review

We monitor for disparate impact across demographic cohorts and apply fairness evaluation to all behavioral signal assessments.

Human Oversight

No automated decisions are produced. All findings are reviewed by qualified personnel before being reported to institutions.

No Consumer Reporting

GoaKarma does not produce consumer credit reports, report to credit bureaus, or communicate findings directly to applicants.

Compliance Awareness

Our evaluation framework is designed with awareness of applicable credit regulations including ECOA, FCRA, and international equivalents.

Signal Taxonomy

Behavioral Indicators Under Evaluation

The following signal categories represent the behavioral dimensions currently under observational study. Each category is evaluated independently for predictive relevance, stability, and demographic fairness before inclusion in any composite assessment.

Signal 01
Payment Consistency
Regularity and timeliness of recurring financial obligations across digital payment channels.
Signal 02
Account Stability
Duration, continuity, and balance patterns within active financial accounts over time.
Signal 03
Digital Financial Activity
Breadth and frequency of engagement across mobile payments, transfers, and digital commerce.
Signal 04
Transaction Patterns
Structural regularity and categorization of spending behavior across observable transaction data.
Signal 05
Community Participation Signals
Peer financial interactions and community-based economic activity as trust indicators.
These indicators are evaluated only for research purposes and are not used to make credit decisions.
Research Prototype

Test the GoaKarma Behavioral Credit Pilot

A voluntary research application used to evaluate behavioral financial signals in thin-file populations.

A limited pilot version of the GoaKarma mobile application is available for research testing. The application collects voluntary behavioral financial signals to evaluate whether they correlate with creditworthiness in thin-file populations.

Collects voluntary financial behavior signals from consenting participants
Supports research into alternative credit indicators for underserved populations
Contributes to anonymized cohort analysis for institutional evaluation

Participation is open to researchers, partner institutions, and eligible individuals who wish to contribute to the study. No credit decisions are made by the application.

Download Test App (iOS) Download Test App (Android)
MVP Pilot Disclosure This application is an experimental research prototype (MVP).
  • The application does not provide loans
  • The application does not generate credit scores
  • Participation is voluntary
  • Data collected is used solely for research evaluation
GoaKarma does not make lending decisions.
GoaKarma · Research Pilot
Behavioral Readiness 742 Credit Readiness Index
Payment Consistency94%
Financial Activity81%
Digital Discipline88%
Community Trust76%
Research Prototype · Not a credit score
Sunset over Goa coastline
Global Human Impact

Credit invisibility is not a personal failure.
It is a structural one.

1.4 billion adults worldwide participate actively in economic life — paying rent, running small businesses, sending remittances — yet remain outside the formal credit system. Not because they are unreliable. Because no system has been built to see them.

01

Economic Mobility

Access to credit is the primary lever by which individuals move from subsistence to stability. A first loan to purchase tools, stock inventory, or cover a medical gap can permanently alter a family's economic trajectory.

02

Correcting Systemic Bias

Traditional credit scoring systems were designed around formal employment and banking relationships. They structurally disadvantage migrant workers, informal traders, and communities that operate outside Western financial architectures.

03

Dignity of Participation

Being seen by a financial institution — having one's responsible behavior recognized and rewarded — is itself a form of social inclusion. GoaKarma's research seeks to quantify what responsible behavior looks like beyond the bureau.

"The objective is financial inclusion. Many responsible borrowers remain excluded not because they are high risk, but because traditional credit systems fail to capture their financial behavior."
Golan Z. Marom
Founder & CEO
Leadership

Golan Z. Marom

Founder & Chief Executive Officer

Golan Z. Marom is a scientist and entrepreneur operating at the intersection of Artificial Intelligence, Machine Learning, Data Science, and Blockchain-enabled trust systems. He is recognized for translating advanced technical research into institution-grade financial platforms trusted by enterprises and regulators.

Research & Thought Leadership

Accepted presenter at the 2026 BIG.AI @ MIT Conference (MIT Initiative on the Digital Economy, April 2–3, 2026). Research: "A Privacy-Preserving, Globally Deployable Credit & Financial Discipline Scoring System" — a consent-first, compliance-by-design credit architecture combining traditional credit indicators with globally portable, privacy-preserving financial discipline signals.

Harvard University — Master's Degree in Design & Applied Arts
MIT (IDSS) — Data Science & Machine Learning
Advanced Specialization — Blockchain: Disruptive Technology
About GoaKarma

Mission

GoaKarma exists to study whether responsible use of behavioral financial data can help lending institutions make better decisions about applicants who are invisible to traditional underwriting.

"The goal is not to replace credit discipline — it is to ask whether we are using all available, responsible information to extend it."

Across the world, billions of individuals actively engage in financial transactions, manage digital payments, and demonstrate consistent financial behavior — yet remain unscored or underscored by formal credit systems. This is a structural problem for lenders and borrowers alike.

GoaKarma's research initiative approaches this problem with the rigor, caution, and institutional respect it requires.

Contact
General InquiriesHello@Goakarma.com
Developer & TechnicalDev@Goakarma.com