Finzy approached us for design and development of a robust, scalable online platform (a web application) that can enable peer to peer lending in India, that can provide lenders an option to manually choose borrowers and invest, or avail the convenience of the platform intelligently auto-investing on behalf of the lender for optimized risks & returns.
Finzy would allow borrowers to apply for loan without collaterals, transact hassle free and get swifter turnaround on loan processing. Lenders can evaluate borrower’s credit score, take calculated risks and exploit lending opportunities. The platform would enable borrowers (individuals) to sign up, share details (like PAN, Aadhaar number) and can upload documents for eKYC, for profile activation and approvals. Lenders can sign up similarly and upload ITR documents for profile activation and approvals.
Finzy also required an intelligent auto-invest algorithm, that splits lender’s investment to fulfil multiple loans (multiple borrowers) in a way that risk is moderated. The platform would analyse borrower’s profile and algorithmically arrive at a “Bridge score” which a lender can view to manually invest in borrowers of their choice or go with the platform recommended mix with moderated risk. Borrowers with low scores would generally have to pay a higher rate of interest, than ones with higher scores.
The borrower name would be masked until the point, lenders decide to transact and every loan would be digi-signed by both parties before transacting. Based on the borrower’s consistent repayment and lenders rating of the borrower, the platform also intends to help improve borrower’s CIBIL scores eventually. The platform was needed to be architected and built in a way it is blockchain-ready by means of which information flow, signing of contracts and transactions can be made transparent to all parties involved.
Codewave architected, designed and developed Finzy.com with extensive features and capabilities for lenders & borrowers to transact; along with an admin interface for the finzy team where interventions were needed. Borrower features included Aadhar verification, eKYC, profile activation, preliminary credit decisioning (Bridge score evaluation) and listing, enabling the profile to be discovered by lenders. Lenders money flows through a NODAL account, and finally reaches the borrower’s account. Both parties can easily view transaction history, an EMI calendar and are notified before an EMI is due. We also developed an admin app with features to monitor transactions, signups and other key performance indicators like pending transfers, new loan applications. Admin is also given an ability to announce new compliances, updates terms and conditions, update interest rate for borrower categories A,B,C. Admin can also blacklist borrowers and lenders.
We recommended MEAN stack for the technology (Read scaling with MEAN), Angular for frontend web development, Express Js for APIs, NodeJS and MongoDB for the backend. On the server infrastructure side, we recommended Multiple Amazon EC2 instances running nginx with elastic load balancing configured for high availability and vertical scaling of applications. For database we recommended MongoDB Atlas for secure MongoDB deployment and also Atlas follows a clustered architecture with a minimum 3 replica sets to ensure high availability of the database and efficient disaster recovery. We chose One signal (used by companies like uber) for push notifications (on the web).