ICICI Appathon 2017
This document proposes a personalized customer experience app for ICICI Bank that addresses key challenges in the banking industry like lack of personalization, non-uniform experiences, and redundant data gathering. The app would provide a one-stop shop with simplified processes leveraging image processing and analytics. It would have an intuitive user interface built using technologies like Ionic, Cordova, Node.js, Python and OpenCV. The app would provide personalized widgets and recommendations for financial products like loans, insurance and investments based on individual customer needs and behavior patterns identified through machine learning. It would also allow for digitized KYC compliance through image uploading and validation. The goal is to improve customer experience, increase cross-selling opportunities, and deliver more value through
ICICI Appathon 2017
- 1. Amit Arora I Uddalak Mandal A Journey Towards Personalized Experience Team Wolverine 1
- 2. Key Challenges Related To Customer Experience Banking Industry Lack of Personalization Paper Driven Processes ( KYC Compliance) Non Uniform Experience ( 3 Different ICICI APPs) Transactional Relationship vs. Trusted Advisor Redundant Data Gathering Non Integrated Services Tremendous Opportunity To Overcome the Challenges via Technology 2
- 3. One Stop Shop Simplified Experience & Process ,Image Processing and Analytics Scalable Technology Stack ( Open Source ) Intuitive User Experience Image pattern identification Simplified Process Ionic 2 Framework Cordova Plugin Node JS ( Platform Independent ) Python | OpenCV Wealth Management - Investments - Banking Insurance -P&C -Life Financial Needs -Home Loan -Auto Loan Solution Overview #ICICIAppathon Seamless Integration ICICI API - getKYC , AddKYC , Behavior Score UIDAI API - Aadhar Card Verification Image API - Custom service 3
- 4. Customer X Auto Loan Widget Customer Y Home Loan Widget Intelligent APP : Personalized Widgets Based on an Individual Needs Use Case 1: Personalized Customer 360 View #ICICIAppathon 4
- 5. Recommendation of Widgets Based on Machine Learning and Analytics Recommendation through Machine Learning #ICICIAppathon 5 Training Dataset Parameters Home Loan Auto Loan Credit card Mutual fund Life Insurance BYFNRNPYSAY Y Y Y Y Y BYFNRNPYSAY Y Y Y Y Y BYFNRNPYSAY Y Y Y Y Y BNFNRYPYSAY N N N Y Y BNFNRYPYSAY N N N N Y Step 1 : Training Data Step 2 : Creating Model using Algorithm RNN (RecurrentNeural Network ) Step 3: Predicting Recommendations
- 6. Recommendation of Widgets Based on Machine Learning and Analytics Use Case : Machine Learning Recommendation ( Admin View ) #ICICIAppathon 6 *Real Time Behavior Score using API , Rest parameters are for illustration purpose
- 7. Simplified Process : Leveraging Image Processing and Analytics Use Case : Digitized KYC Compliance ( S3V : Select | Scan | Submit | View) #ICICIAppathon 7
- 8. Delivering Value Through Technology #ICICIAppathon Increase Cross & Up Sell Opportunities ECONOMIC ENVIORNMENTEXPERIENCE Revenue Customer Loyalty Increase Net Promoter Score Giving Back Reduce Paper Consumption 8