Skip to content

9.2 Headcount by AUM

The headcounts below are indicative, not prescriptive. Adjust to the specific operating mix (more anchor programmes need more partner ops; more own-channel needs more sales).

Total: ~20 – 28 people including founders.

FunctionHeadcount
ExecutiveCEO + 1
Sales / channels2 – 3
CreditHead + 2 – 3 analysts
OperationsHead + 3 – 5 (KYC, disbursement, recon, customer support combined)
Collections1 – 2 (in-house) + outsourced agents
Engineering6 – 10 (platform, data, integrations)
FinanceHead + 1 – 2
Compliance + CSCompliance officer + CS
RiskCRO (often combined with credit head at this stage)
InfoSec1 (could be fractional + vendor SOC)
Vendor / partner mgmt1

Total: ~50 – 75 people.

FunctionHeadcount
ExecutiveCEO + COO + 1
Sales / channels5 – 8 (split by partner, DSA, direct, CA)
CreditHead + 5 – 8 analysts + risk analyst
OperationsHead + 8 – 12 (KYC + disbursement + recon + LMS ops + customer support)
Collections5 – 8 in-house + outsourced
Engineering15 – 22 (split by module teams)
Data / Analytics2 – 3
FinanceHead + 3 – 5
ComplianceCompliance Officer + 1 – 2
RiskCRO + 1 – 2 (separated from credit by now)
InfoSec2 – 3 (incl. SOC analyst)
Vendor / partner mgmt2 – 3
Internal Audit1

Total: ~120 – 180 people.

FunctionHeadcount
ExecutiveCEO + COO + Heads
Sales / channels15 – 25 (multi-channel)
CreditHead + 12 – 18 (analysts + senior + segment-specific)
RiskCRO + 4 – 6 (model, monitoring, ALM)
OperationsHead + 25 – 35 (departments)
Collections12 – 20 in-house + scale outsourced
Engineering35 – 55 (platform teams)
Data + ML8 – 12
FinanceHead + 8 – 12
Compliance4 – 6
InfoSec5 – 8
Vendor / partner mgmt5 – 8
Internal Audit2 – 3
Legal2 – 3
  • Engineering scales superlinearly with platform complexity, not book size. A ₹300 Cr book with 3 co-lending partners needs more engineering than one with 1 partner.
  • Collections scales with NPA pipeline, not book size alone. A book with 3% NPA needs different staffing than 1% NPA.
  • Sales/channels scales with origination strategy. Heavy CA / partner / embedded distribution scales slower in headcount than direct DSA distribution.
  • Compliance and InfoSec must scale with regulatory complexity, not strict-line book size. Going NBFC-ML or onboarding additional partners adds non-linear obligations.

The recurring trap: hiring sales / DSA aggressively before underwriting / collections capacity is built. The result is a surge of poor-quality applications, declines (or worse, approvals that go bad), DSA frustration, and reputation damage. Hire credit and collections ahead of sales, not behind.