Skip to content
by Rakesh Waghela

B2B Lending India

A working spec for B2B lending in India — origination, underwriting, disbursement, servicing, collections. Built for SME working-capital, CA/Tally/GST origination, repeat short-tenure loans, and co-lending with banks/NBFCs.
Start with the recommendation Reading paths by role GitHub

What this site is

Section titled “What this site is”

A working, India-specific, implementation-grade blueprint for a B2B lending platform built around four anchor models:

  1. Own-book / managed NBFC lending — ₹30 crore initial participation.
  2. Co-lending with banks and NBFCs — ₹70 crore+ partner participation.
  3. LSP and SaaS-platform fee model — annual platform fees, setup fees, servicing fees.
  4. SME working-capital, repeat short-tenure — ₹20 lakh to ₹50 lakh tickets, 60–180 days.

Origination wedge: CA networks, Tally / Zoho Books / GST data, distributor and vendor ecosystems. Not consumer, not BNPL, not app-only.

How the spec is organised

Section titled “How the spec is organised”
1. Business models Own-book, co-lending, LSP, marketplace, SCF, anchor-led, invoice discounting, white-label, SaaS — money flow, who owns risk, regulatory shape, unit economics.
2. Regulatory framework RBI Digital Lending, DLG/FLDG, co-lending, KYC/CKYC, AA, asset classification, FPC, IT/cyber, outsourcing, DPDP, compliance calendar.
3. Product module map A–P — acquisition, application, KYC/KYB, ingestion, underwriting, co-lending, docs, disbursement, LMS, collections, monitoring, accounting, reporting, admin, analytics.
4. Integrations and vendors Bureaus, KYC/KYB, AA, BSA, GST, accounting software, eSign/eStamp, NACH/UPI, comms, FI, fraud, banking — with Indian vendor recommendations.
5. Architecture High-level services, data model, event model, workflow engine, rule engine, security/IAM, audit, data platform.
6. Underwriting Rule library, bureau/GST/bank-statement/cash-flow/Tally/fraud rules, exposure-pricing-tenure grids, MVP vs advanced scorecard.
7. Co-lending deep dive Operating model, allocation engine, APIs, settlement, reconciliation, DLG monitoring.
8. Vendors — build vs buy Lentra, M2P Finflux, Finezza, CloudBankin, Roopya, Decimal, Biz2X, Perfios, Yubi, CredAble, Veefin — what to buy, what to build.
9. Operations and team Org design, headcount by AUM, tools, KPIs, failure modes.
10. Unit economics Three scenarios — ₹30Cr own, ₹100Cr blended, ₹330Cr blended. Sensitivity to NPA, fees, cost of funds.
11. Risk register 25 risks with controls, early-warning signals, system features.
12. MVP roadmap Phases 0–7 from compliance design to scale automation.
13. Implementation backlog 21 engineering epics, user stories, APIs, edge cases, acceptance criteria.
14. Tech stack Java/Spring Boot/PostgreSQL bias — services, data, workflow, rules, infra, security.
15. Final recommendation The blunt answer — wedge, model, 90-day plan, 12-month roadmap, fatal mistakes.
16. Client diligence and suitability Pre-application diligence, borrower-type protocols, promoter background, industry compliance, customer/supplier verification, physical verification, continuous monitoring, suitability, recovery-readiness.
17. Recovery and resolution Strategy by bucket, legal pipeline (notice / arbitration / civil / DRT / IBC), SARFAESI, settlement / OTS, NBFC-ARC transfer, recovery analytics.
18. LOS + LMS deep-dive Bit-level spec — state machines, every call and operation, interest conventions, schedule algorithms, daily batch, charge ledger, waterfall math, co-lending share accounting, bureau format, reconciliation, RBI returns, edge cases.

Who should read which section

Section titled “Who should read which section”

Each role has a recommended path through the spec — see Reading paths.

Author

Section titled “Author”

By Rakesh Waghela — Tech and B2B Lending Solutions Architect. LinkedIn · X · Book a consultation