India feeds the world.
Nobody funds the kitchen.
The AI-first lending platform for India's ₹5.7 lakh crore HORECA industry — built on a live, working underwriting engine you can test today.
This is not a concept deck.
Everything here is running in production, right now. You're welcome to open it on your phone while we talk.
6-factor HORECA credit model scores any outlet in under a second and writes a full AI risk memo.
Applications, loan book, virtual-account collections engine, DPD buckets, portfolio analytics. 17-table production database.
Credit policy, collection & recovery policy (Fair Practices Code), bureau reporting framework — written, reviewed, board-ready.
Kitchens run on credit nobody gives them.
A restaurant earning ₹7 lakh a month — daily cash flow, real customers, real margins — walks into a bank and gets rejected. Why?
30–40% of HORECA revenue is cash & UPI that annual filings under-report. Banks score the paperwork, not the business.
A dhaba and a steel trader get the same scorecard. Margin structures, seasonality, footfall — all invisible to NBFC models.
HORECA earns daily but repays monthly — one bad week breaks the EMI. Product design itself manufactures defaults.
3rd-largest food market on Earth by 2028.
Working capital + expansion credit across 14M+ outlets (~12% of GMV)
Customer-facing outlets in top 20 cities with digital payment trails
~26,000 active loans at ₹6.5L avg — just 0.2% of outlets
Built around how a kitchen actually earns.
- 6-factor HORECA scoring — margin, growth, sales, location, ambience, bureau
- Reconciles bank + aggregator data; scores the agreement, not the claim
- Rasoi IQ writes a full, auditable credit memo for every decision
- A regulated virtual account sits in the borrower's settlement path
- Swiggy/Zomato payouts route in; EMI auto-deducted before the borrower is paid
- Remainder swept to the restaurant — repayment is structural, not behavioural
- 5-bucket framework (Clean → NPA) under RBI Fair Practices Code
- Escalation proportional to DPD — relationship first, recovery second
- Clean payers earn better terms + cross-sell after 6 months
We don't chase repayment. We collect at the source.
Every competitor debits the borrower's account and hopes the balance is there. We take our EMI before the money ever reaches the borrower — powered by regulated virtual-account escrow rails, the same architecture Mahindra Finance, IIFL and Oxyzo already run on.
A regulated virtual account opened at loan origination. It's theirs — but it sits in the settlement path.
Swiggy & Zomato settlements land here — where 30–60% of a modern outlet's revenue already flows.
Weekly EMI auto-deducted at source; the remainder swept to the restaurant, same day.
The money is collected before it can be spent. A borrower can't forget, go overdrawn, or deprioritise us.
No field collection, no bounce cycles for the performing book. The rails do the work.
At-source collection cuts expected loss on settled revenue — so we approve outlets a bank rejects, below informal rates.
We don't score what a restaurant claims. We score what independent sources prove.
Any lender can read a bank statement. Our engine reconciles independent data sources and scores the agreement between them — confidence from corroboration, not from documents a borrower could fabricate.
Bank statements + Swiggy/Zomato payouts matched transaction-by-transaction. Claimed sales must show up as money that landed.
How well independent sources agree. TRUSTED / REVIEW / FLAGGED — computed before we ever score creditworthiness.
Round-number clusters, circular transfers, manufactured consistency force the riskiest band. Built to catch the file that's too clean.
Before a document is uploaded: ratings, price band, location, nearby demand (colleges, tech parks). Real-world presence becomes a credit signal.
Let's underwrite a real restaurant. Right now.
We already have the rails to reach every kitchen in India.
Most lenders spend years and crores acquiring SME borrowers. We start with distribution built in — three channels, each a warm pipe to the outlets we underwrite.
Three of our founders built and exited Tipplr on the ONDC network. Rasoi inherits that live integration — programmatic reach to MSME restaurants nationwide. Not a partnership we hope to sign; a rail our own team built.
The National Restaurant Association of India is the collective voice of the sector. A channel here puts Rasoi in front of organised HORECA demand and the operators who set norms.
Petpooja (150K+ outlets), UrbanPiper (45K+) and peers sit on daily transaction data. Embedded lending turns every terminal into an origination point. Revenue-share, near-zero CAC.
Five unfair advantages. None of them copyable.
6-factor HORECA model with CIBIL as the smallest weight. That inversion approves outlets banks reject — at lower NPA.
Our regulated virtual account intercepts Swiggy/Zomato settlements; EMI deducted first. Structurally impossible to replicate on debit-time architecture.
150K+ POS outlets plus a founder-built ONDC integration. Competitor field-sales CAC: ₹8–15K/loan. Ours: near zero on repeat.
Every rupee of a borrower's aggregator revenue flows through the account we control — live revenue data feeding the next credit cycle. Monthly lenders see 12 payments a year and hope.
Credit policy, Collection & Recovery Policy (Fair Practices Code), bureau framework — written and board-ready. Most applicants spend 6–9 months here.
Nobody underwrites a kitchen. Until now.
| Oxyzo | LendingKart | FlexiLoans | Indifi | NeoGrowth | RASOI | |
|---|---|---|---|---|---|---|
| HORECA-specific scoring | ✗ | ✗ | ✗ | Partial | ✗ | ✓ 6-factor |
| Field intelligence (FSA) | ✗ | ✗ | ✗ | ✗ | ✗ | ✓ Built-in |
| Collection at source | ✗ | ✗ | ✗ | ✗ | Weekly (POS) | ✓ Virtual account |
| AI credit memo per loan | ✗ | ✗ | ✗ | ✗ | ✗ | ✓ Rasoi IQ |
| Borrower-friendly bureau | ✗ | ✗ | ✗ | ✗ | ✗ | ✓ Monthly-cum. |
| Typical rate (SME) | 12%+ | 18%+ | 19%+ | Varies | 20%+ | 18–27% risk-based |
Every loan we make, makes the next loan smarter.
FSA captures ambience, location, category — data no bureau sells.
Every settlement flows through the account we control — we see real revenue, not an inferred proxy.
Outcomes feed rl_training_data — the RL model is already architected in the production schema.
Better approvals at the same risk → lower NPA → cheaper debt capital.
Unit economics of a single loan.
Illustrative · Bucket C · ₹7L · 24 months · 19% p.a. + 2.5% PF
24 months · prove-out · ₹7L cap
Repeat borrower · known payment history · uncapped
Rate discounts via clean record · CAC ≈ ₹0 on repeats
Two roads to the licence. Both fully costed.
RBI mandates ₹10 Cr minimum Net Owned Fund either way (Sec 45-IA, RBI Act) — it anchors this raise.
- Buy a clean, dormant NBFC-ICC shell with valid CoR
- RBI prior approval — 26%+ stake / change of control + fit-&-proper
- 30-day notice, then capital infusion to ₹10 Cr+ NOF
- Premium over book: ₹50L–1.5Cr by vintage
- Incorporate, capitalise ₹10 Cr NOF ab initio
- COSMOS application — business plan, 5-yr projections, fit-&-proper
- Clean diligence trail — no legacy book, no inherited risk
- Our policy stack is already application-grade
Serial entrepreneurs who've built, scaled, and exited.
Three of us built and exited Tipplr together (acquired by GoKhana, Jan 2026) — a restaurant-commerce business on ONDC. We're now building the lending layer for the market we already know cold.
Serial entrepreneur — "I build, scale, and exit marketplaces." Co-founded Tipplr (acquired by GoKhana, Jan 2026). 20+ years across marketplaces and enterprise technology. Leads product, technology, and fundraising.
Founder Director of India's first credit-advocacy company. Works with regulators and 60+ lenders; author of "Why is my Credit Report Screwed up" (Network18); columns in Mint, Femina, Outlook Money; TV credit expert. Leads credit policy & regulator relationships.
Co-founded Tipplr. 20+ years across enterprise IT, data platforms, and startups — Engineering, Product & Data across Retail, FMCG, Banking, Insurance, FoodTech. Hands-on in AI/ML and ONDC. IICA-Certified Independent Director. Leads engineering & the AI platform.
Founder of Petoo, supplying ready-to-cook products to 300+ restaurants monthly across North-East India. IIM-Bangalore; 18+ years in banking & telecom (ICICI, Vodafone Idea) in credit & prepaid products. Already sells to the outlets we lend to.
Co-founded Tipplr; Founder of Vanbey Services. Serial startup mentor and early-stage investor with 30+ years of marketing & business development across Asia. Leads go-to-market and distribution.
Domain + credit + technology + distribution + operating experience — every leg of a HORECA lending business covered by someone who's actually done it.
The kitchen always pays.
Now someone finally lends to it.
Live AI underwriting + full lending stack — demo in your hands today
₹10 Cr NOF priced in · both licence paths costed · policies board-ready
We control the account every rupee of aggregator revenue flows through
~80% of funds become NOF + loan book, not burn