Lending, kept like a ledger:
precise, priced, explained.
Gradexa pairs a deterministic Credit Engine with a complete lending Suite — CRM, LOS, and LMS on one data model. Decide, price, disburse, and service, with every decision you can defend.
A decisioning API and the Suite it powers.
Use the Credit Engine on its own via API, adopt the full Suite, or run both — the same engine can power your own lending product.
Credit Engine
A deterministic decisioning brain. Send it the bureau report you pulled with your own bureau membership; get back a risk grade, a priced offer, and the exact reasons — every time. 10 risk grades, each with pricing bands your credit team configures. Rates are set by the lender, not by Gradexa.
- 01Two-stage: eligibility knockouts → 10-grade pricing
- 02Explainable: fired rules + score contributions
- 03Your rules, no code — configuration, not deployment
The Suite
Full-stack lending infrastructure on one data model. Single-repayment and EMI loan products — EMI over 3–12 months on a reducing balance, with an amortization schedule fixed at disbursal, part-payment with schedule recomputation, pre-closure, and per-installment DPD.
- 01CRM: lead → KYC → borrower 360
- 02LOS: application → product selection → decision → e-sign → disbursal
- 03LMS: servicing both products — single-repayment payoff and per-installment EMI schedules
From a lead to a serviced loan, end to end.
The Suite carries an application into the Credit Engine and back out as a decision — then straight into servicing. Nothing is re-keyed between stages.
- 01CRMLead · KYC · borrower 360
- 02LOSApplication · product selection
- 03Credit EngineKnockouts → 10-grade pricing
- 04DecisionGrade · offer · fired rules
- 05LMSDisburse · service · collect
One data model end to end — no re-keying, no lossy hand-offs. Every decision is traceable back to the rules that fired.
Built for lenders who have to explain themselves.
Regulators, credit committees, and borrowers all ask 'why?'. The platform is designed so you always have the answer.
Explainable by design
Every decision returns the rules that fired and each factor's contribution to the score. No black box — defensible to a credit committee and an auditor.
Your rules, no code
Eligibility knockouts and pricing bands are configuration, not a deployment. Credit teams change policy without waiting on an engineering release.
One data model
CRM, LOS, and LMS share a single borrower and loan model. A lead becomes a borrower becomes a serviced loan — no re-keying, no reconciliation drift.
Tenant data isolation
Each tenant's data is isolated. The platform is multi-tenant by design so your book stays yours.
An LMS that keeps your books the way the RBI expects them.
Most lending stacks stop at disbursal. The ledger underneath ours classifies, suspends and appropriates every day, so month-end is a report, not a reconstruction.
- 01IRAC classification: Standard → SMA → NPA at 90+ DPD, automatic
- 02Suspended interest on NPA accounts, realised only on cash
- 03Configurable payment appropriation, per product and for NPA
- 04Custom DPD buckets per product; single-repayment and EMI on one ledger
Straight talk on what's built
Gradexa is in active development. The decisioning logic, pricing grades, explainability, single-repayment and EMI loan products (EMI over 3–12 months), and the shared CRM/LOS/LMS data model are the core of the platform today. The engine makes no bureau calls: you pull the report with your own bureau membership and send it, and the engine parses and normalises it (CIBIL XML, CRIF JSON, Experian fixed-width, Equifax JSON). Direct bureau pull, soft or hard, is on the roadmap. NACH mandates are partial: core logic built, not production-ready. We have no production lenders yet, and we don't publish customer logos or latency figures we haven't earned.
See it decide on a real bureau report.
Request a demo and we'll walk you through the Credit Engine and the Suite on your scenarios. Honest about what's built today, what's partial, and what's on the roadmap.