AI-native credit decision engine for India's MSMEs
A full-stack credit decision engine for India's MSMEs — turns a GSTIN into a risk score, probability of default, loan recommendation, and explainable decision summary in under 5 minutes.
54 million businesses are invisible to lenders
India's MSMEs contribute 30% of GDP and employ 25 crore people. Yet the majority have never accessed formal credit — not because they're risky, but because traditional bureaus like CIBIL have no data on them.
NBFCs want to lend to these borrowers. But without a credit score, underwriting is slow, manual, and inconsistent. Loan officers spend days reviewing documents that could be analyzed in minutes.
Manual underwriting takes days
Credit teams manually review GST returns and bank statements — slow, expensive, and inconsistent across officers.
CIBIL returns "No History"
54 million MSMEs have no bureau record. NBFCs either reject them or take on unquantified risk.
₹25 lakh crore credit gap
Formal credit demand from MSMEs far outstrips supply. The gap grows every year.
Credit scoring in minutes, not days
MAARU integrates with your lending workflow via a single API call. Send a GSTIN — get back a full credit decision: score, default probability, loan recommendation, and a plain-English explanation.
GSTIN input
NBFC submits a borrower's GSTIN via API or dashboard. One call — no documents, no manual data entry required.
GST data fetched
Registration, filing history, and business status pulled directly from GSTN in real-time. Account Aggregator integration on our V2 roadmap.
Scoring + PD engine
Weighted scorecard across 5 GST-derived factors, followed by a logistic PD model with penalty flags and loan sizing recommendation.
AI narrative + PDF report
Claude writes a plain-English credit narrative. A branded PDF is auto-generated with full breakdown, decision summary, risk flags, and loan recommendation.
Dashboard for underwriters
Credit teams review decisions, drill into score breakdowns, and download reports — no code required. API plugs directly into your LOS.
Built for how NBFCs actually lend
Probability of Default Engine
Logistic PD model calibrated to Indian MSME risk profiles. Returns PD% with portfolio benchmarks, peer percentiles, and a 5-tier risk band.
Penalty-based risk scoring
Stacked penalty flags — inactive GST, litigation, negative news, filing delays — each add specific PD deltas with full transparency on impact.
Loan recommendation engine
Returns max loan amount, interest rate range, tenure, and collateral requirement — sized to the borrower's GST-derived revenue and risk profile.
Explainable decisions
Every decision comes with a risk summary, decision rationale, structural vs correctable risk classification, and trend insight — not just a number.
API-first, PDF-ready
Plug into your LOS via REST API. Every score auto-generates a branded PDF credit report with full breakdown, decision summary, and AI narrative.
Account AggregatorComing Soon
Bank-statement scoring via RBI's Account Aggregator framework for consent-based data flows — on our V2 roadmap.
Multi-language supportComing Soon
Consent flows and borrower communications in Hindi, Tamil, Telugu, Kannada and more — rolling out in a future release.
Lend faster. Lend smarter. Lend more.
- Get PD%, not just a score — price risk with precision
- Explainable decisions your credit committee can defend
- Score borrowers CIBIL marks as "No History"
- Penalty flags surface hidden risks before disbursement
- Loan sizing recommendations calibrated per borrower
- Expand into Tier 2/3 markets with confidence
Credit should be a right, not a privilege
India's 63 million MSMEs are the backbone of the economy — employing 25 crore people and contributing 30% of GDP. Yet the majority have never accessed formal credit, not because they're risky, but because they're invisible to traditional systems.
MAARU exists to change that. By using the financial data these businesses already generate — GST filings, UPI transactions — we make the invisible visible. Every small business owner who gets a fair shot at credit is a win for India.
Abhishek Sharma
Data Engineer · MS Computer Science, SUNY Binghamton · Delhi, India
Built after watching creditworthy MSMEs get rejected
After watching NBFCs reject creditworthy MSMEs because CIBIL returned 'No History', Abhishek built MAARU — a scoring engine that reads the financial data these businesses already generate: their GST filings. India's MSME credit problem is fundamentally a data problem, not a lending problem.
The financial signals to assess creditworthiness already exist in GST filings and UPI transactions. They just need the right engine to interpret them.
MAARU is his answer to that — a purpose-built AI scoring engine that turns existing financial data into credit decisions, giving 54 million MSMEs their first real shot at formal credit.
abhishek@maaru.aiPartner with MAARU
We're onboarding our first NBFC pilot partners. If you're a lender looking to expand MSME credit access, let's talk.
Free 30-day pilot for qualifying NBFCs · Paid plans from ₹500/report
Apply for a Free Pilot
We score 50 of your live applications at no cost. You keep the reports. Tell us about your portfolio and we'll be in touch within 24 hours.
Or email us directly at hello@maaru.ai