Tappo.ai — Digital Micro-Lending Platform

Traditional lending is slow, paperwork-heavy and out of reach for many people — applying means branches, forms and long waits, and lenders struggle to assess risk for customers without a long credit history. Tappo.ai set out to deliver micro-loans and buy-now-pay-later financing entirely through a mobile app: instant onboarding, fast eligibility decisions, and simple repayments. Doing this well meant solving identity verification, risk assessment, loan management and repayments together — in a system reliable enough to handle money and scale with demand.

We built the Tappo.ai platform end to end: a native mobile app for customers, a web admin panel for operators, a marketing website, and the backend services behind them. Customers onboard through a guided flow with identity verification and document capture, check their eligibility, take out installment loans or buy-now-pay-later purchases from partner merchants, and manage repayments — all from their phone, secured with a PIN and biometrics. Loan applications are assessed automatically by an AI credit-scoring engine, then routed to operators who review, approve and manage them through the admin panel, alongside borrower management, merchant management and financial reporting. The backend is split into focused services — a core lending engine, a merchant inventory service and a notification service — with payments, KYC and notifications integrated throughout. The app is fully multilingual, including right-to-left support.

Tappo.ai lets customers go from sign-up to a loan decision in minutes, entirely on their phone, with identity verification, AI-based scoring and repayments built into one smooth flow. Operators get full control over the loan lifecycle — review, approval, collections and reporting — from a single admin panel, while the buy-now-pay-later integration extends lending into real purchases at partner merchants. The service-based backend keeps the different parts of the system independent and scalable, and the multilingual app makes the product accessible across its markets.

• Fully Mobile Lending: Customers onboard, get scored, borrow and repay entirely in a native app.
• AI Credit Scoring: Applications are assessed automatically for faster, data-driven decisions.
• Secure Onboarding: Identity verification, document capture, PIN and biometric protection.
• Buy Now, Pay Later: Installment financing for purchases at partner merchants.
• Operator Admin Panel: Role-based review, approval, collections, merchant management and reporting.
• Multilingual & Accessible: Full localisation including right-to-left language support.

• React Native: The native iOS and Android customer app, with biometrics and document capture.
• Django & Django REST: The core lending engine — loans, customers, KYC, repayments and transactions.
• FastAPI: Lightweight services for merchant inventory and notifications.
• PostgreSQL: Core data store across the backend services.
• RabbitMQ & Celery: Asynchronous processing for scoring, notifications and background work.
• React (Vite) with Ant Design: The operator admin panel.
• Firebase & email/SMS providers: Push notifications and transactional messaging.
• Docker: Containerised, reproducible deployment of every service.

We designed Tappo.ai around the customer journey — onboard, verify, get a decision, borrow, repay — and built the backend as focused services so each concern could evolve independently: a Django core for the lending lifecycle, plus lightweight FastAPI services for merchant inventory and notifications, communicating asynchronously through a message queue. Identity verification and document capture were built into the mobile onboarding flow, and loan applications were wired to an AI credit-scoring engine for fast, automated assessment. We built the operator admin panel for the full loan lifecycle — review, approval, collections and reporting — and added the buy-now-pay-later merchant flow on top. The native app was built for security and accessibility, with PIN and biometric protection and full multilingual support, and everything was containerised for reliable deployment.