Prayas is a analysis tool that:

• Predicts bad loan takers by analysing previous loan records.
• Provides an alternate means to credit score (CIBIL) by doing a risk analysis of the loan taker.
• Provides means to take care of NPAs (Non performing assets)

Problem

• Bad loans in banks crimp profits, chokes credit flow.

• Non-performing assets or NPAs of banks have risen to about 17% of total loans.

Solution

• Identify risky loan takers beforehand by predicting behaviour based on previous financial activity.
• Designing customer strategies by admin by detailed analysis of the loan seekers.

Methodology

1. Study the user's financial history.
2. Classify stats into behavioural patterns.
3. Admin views detailed analysis of user data.
4. Design customer strategies based on the usage patterns.
5. Preemptively decide if loan could become an NPA
6. Recommend actions to bank to protect the loan

Technology Stack:

• Python 
• Flask
• Scikit
• MongoDB

• Customer, Account and Loan Account APIs
• Microsoft Cognitive API
• Crediwatch API

APIs Used: