CortexData
For NBFCs

The full lending stack for NBFCs scaling past ₹1,000 Cr.

Origination. Management. Co-lending. Securitisation. ML decisioning. One platform, RBI-aligned, India-ready.

CortexData is the lending OS for the NBFC scale-up problem: you've outgrown your original LOS, your co-lending programmes are reconciliation nightmares, and your treasury wants to securitise. Instead of stitching together six vendors, run it on one platform built for India's regulatory reality.

The problems we keep seeing in scaling NBFCs.

If three or more of these are happening at your firm right now, CortexData solves the next year of your roadmap.

Pain

We grew origination 5x in two years on duct-taped LOS workflows. Now scaling another 3x is breaking. Adding products takes months. Underwriter throughput plateaus. Field teams use WhatsApp because the platform is too slow.

What CortexData does

CortexData LOS scales horizontally per service. Adding a product line is a configuration + scorecard change, not a build. The underwriter screen is designed around a 7-minute decision cycle: claim → review → decide → next.

Pain

Every co-lending programme with a bank partner means a new spreadsheet for settlement reconciliation, a new email chain for partner KYC, and a new compliance person checking RBI co-lending caps weekly.

What CortexData does

CortexData Co-Lending implements the RBI Co-Lending Master Direction (Nov 2020) end-to-end. Originator + partner roles, escrow disbursement, daily settlement reconciliation, partner exposure caps, single-customer-interface enforcement.

Pain

Our treasury wants to securitise. Investment bankers quote ₹50 lakhs for 'platform setup' alone. Each issuance is a 4-month bespoke build because no off-the-shelf platform supports the RBI 2021 securitisation framework with PTC issuance + investor onboarding.

What CortexData does

CortexData PTC implements RBI Securitisation Master Direction 2021. Pool eligibility, MRR/MHP, true-sale verification, tranching, cash-flow waterfall, investor onboarding, secondary trading, RBI quarterly reporting — all built-in. Issuance becomes a configuration, not a project.

Pain

RBI returns (IRACP, PSL Form A, DSB) take 3-5 person-days each because the underlying data lives in 6 systems. Reconciling DSB to LMS to bank statement is a manual exercise. Inspectors find inconsistencies. Penalties happen.

What CortexData does

All four returns generate from the same source-of-truth in the LMS. Daily DSB, monthly IRACP, quarterly PSL Form A, quarterly OSS-3 — same data, regulator-ready format, document IDs preserved for audit.

Pain

Our credit committee wants ML scoring. The data scientist built an XGBoost model in a notebook. Productionising it has been 'next quarter' for 3 quarters. Meanwhile underwriters override the rule engine 30% of the time, and we have no audit defensibility for those overrides.

What CortexData does

CortexData ships ML decisioning as a production layer: 5-model fraud ensemble, calibrated PD model, per-decision feature attribution into the audit chain. Trained on YOUR data, on-premise. Underwriter override paths are first-class — every override is captured, attributed, and reviewable.

What changes after CortexData.

  • Underwriter throughput up — 7-minute decision cycle vs. 25+ on legacy
  • Co-lending programmes go from 6-month builds to 3-week launches
  • Securitisation issuance moves from spreadsheet to platform
  • RBI returns generated in seconds, not person-days
  • ML scoring that's trained on YOUR data, on YOUR infrastructure
  • Per-decision audit trail defensible to RBI and the Banking Ombudsman
  • Multi-tenant: one platform serves your wholly-owned subsidiaries
  • API-first: plug into your existing CRM, KYC stack, and notification systems

From one product to four-product co-lending in 6 months.

Bring us your next product launch or your next co-lending programme. We'll model it on CortexData and walk you through the full lifecycle — origination to disbursement to settlement to RBI return.