Bajaj Finance targets 3,000-store facial-recognition rollout by March 2027
Bajaj Finance plans to scale AI-powered cameras from 544 stores to 3,000, using real-time customer identification to tailor offers, improve onboarding and curb fraud. It has captured nearly 10 million images in four months and identified 3.2 million existing customers.
What happened
Bajaj Finance plans to deploy AI facial-recognition cameras across 3,000 Indian stores by March 2027, enabling real-time identification and personalized offers.
Key facts
- Facial-recognition system to expand from 544 stores to 3,000 stores by March 2027
- Nearly 10 million images captured over four months
- 3.2 million existing customers identified
- Identification rate targeted at 70% by March 2027 from about 40% currently
- Consolidated net profit: ₹6,081 crore, up 28% year-on-year
- AUM: ₹5.5 trillion as of 30 June, up 24% year-on-year
- India domestic credit: ₹235 trillion; projected ₹729 trillion by FY36 at 12% annual growth
- Bajaj Finance AUM: ₹5.1 trillion in FY26; projected ₹40 trillion by FY36 at 23% growth
Why this matters
The rollout strengthens the case for partnerships or acquisitions in consent management, edge AI vision, identity orchestration and fraud analytics that can make physical networks data-rich channels.
What to watch
- Identification rate progression from roughly 40% toward the 70% target.
- Incremental loan disbursal, cross-sell conversion and turnaround-time improvement at enabled stores.
- Customer opt-out rates, complaints and any regulatory guidance on facial recognition or biometric data.
- Evidence that facial matches reduce fraud losses or duplicate-account/onboarding attempts.
- Expansion pace beyond 544 stores and whether the company remains on track for 3,000 by March 2027.
- Rival deployments of biometric, computer-vision or app-led in-store personalization.
- Link identified store visitors to a unified CRM journey spanning app, call center, merchant and branch interactions.
- Prioritize pre-approved, low-documentation offers and service resolution rather than broad promotional targeting.
- Build explicit biometric consent, opt-out, data-retention and audit controls before scaling to 3,000 locations.
- Use store-level test-and-learn measurement to compare facial-recognition-driven uplift against conventional lead capture.
- Extend the capability to partner merchant locations, where transaction context can improve lending and cross-sell relevance.