Engineering a Loan Marketplace with Automated Filtering

At Internovo, we identified a bottleneck in our lending operations: bankers were spending hours manually screening loan applicants.

The Vision

Create a marketplace that:

Technical Implementation

Database Design (MySQL)

sql
CREATE TABLE loan_applicants (
  id INT PRIMARY KEY AUTO_INCREMENT,
  credit_score INT,
  income DECIMAL(10,2),
  loan_amount DECIMAL(10,2),
  loan_purpose VARCHAR(100),
  employment_status ENUM('employed', 'self_employed', 'unemployed'),
  created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);

CREATE TABLE lenders (
  id INT PRIMARY KEY AUTO_INCREMENT,
  min_credit_score INT,
  max_loan_amount DECIMAL(10,2),
  interest_rate DECIMAL(5,2),
  preferred_sectors JSON,
  is_active BOOLEAN DEFAULT true
);

Matching Algorithm

The core algorithm considers:

Frontend (Next.js)

Results

Lessons Learned

1. Data Quality Matters

The algorithm is only as good as the data. We spent significant time on:

2. Compliance is Complex

Financial products require:

3. UX Drives Adoption

Bankers are busy. The interface needed to be:

This project reinforced my belief that well-designed internal tools can transform business operations.