Technology Partner

Bank of Baroda Hackathon - 2022

Your Team Name :W Technnovators

Your team bio : Soni Kumari

Bibha Priya
Apurva Patil

Date : 7 Sep 2022

///01
/1
 Call Center Analytics-Why we need to relook?
Call Centers- Frontline of Most Business Strategically Important Future capabilities
• Direct point of contact with customers • Increases customer loyalty and retention • Repository of data
• Great source of information about customer • Enables increase in sales and loss of • AI/ML capability utilization to convert customer’s
perception across products opportunities bitter experience to WOW moment
• Enables to relook processes/products which are • A competitive advantage • Track customer sentiment over time
pain point for customers • Strategizing improvement in customer experience • Enable management to get quick feedback from
• Provide actionable insights • Increasing Brand Value and Market Share customers bypassing organizational hierarchies
• Identifies areas to improve upon customer service • Reduces cost by shared support center • Enable quick decisioning through various
dashboards

WHY BANK OF BARODA NEED TO FOCUS ON CALL CENTER ANALYTICS?

BOB SBI HDFC ICICI Moment of Truth
Deposit Trust
410,123 3,920,201 155,920 106,450 Positive Feedback Negative Feedback
Loans &
Efficient Process Perception of Bank Decreased value of prod-
15
Advances 684,153 2,818,671 138,050 85,900 ucts purchased
improved
13
Increased but did nothing Switched Perception of Bank
NIM 3.14% 3.12% 4% 3.96% Profitability value of 14
main bank worsened but did
28
products nothing
ROE 11.82% 13.92% 16.90% 14.80% Shareholder’s purchased from
58
Confidence bank Purchased
New Product Stopped product
29
• Competition from large Public Sector Banks, 20
Private Banks, Small Finance Banks, Fintechs etc Bought product from an-
23
other bank
• Innovative Deposit and credit products in the
market >85% customers purchased more >70% customers purchased less
• Stringent regulatory guidelines
*3 out of 5 surveyed customers would likely to abandon their favorite brand after a bad
customer service interaction
2 out of 3 customer are unlikely to purchase again from a favorite brand if they received no
solution after a complaint.
Mckinsey Survey 2004-  Retail-banking industry  Belgium, Germany, and
Italy ///01
/2
 Call Centers- Forming Emotional bonds to Create Business Values
Critical for Industries Key Challenges What is important for Customers Trust?
• Financial Services • Increasing volume Quality Products or Services 83%
• Consumer • Complexity of Interactions Fair Prices 71%
• Services • Agility of Service- Due to Pandemic Issues Fixed in a Timely Manner 55%
• Technology, Media & Telecommunication • Providing Right Channel
• Life Sciences & Healthcare • Breaking Business Silos
Helpful Employees 48%
• Energy, Resources & Industrials • Data Analysis for comprehensive view History with the customers 40%
• Government & Public Services of customer’s voice and sentiment. Business Ethics 38%
Honest Messaging 37%
Environmental impact 20%

Call Center Investment- Main Drivers AI/ML Based Call Center Analytics-Key Enablers

85 •
Customer Experience 88 Using AI to foster human connections at machine speed
71 • Drive more personal, intuitive and empathetic customer experience
78
Importance of Service 73 79% of contact center leaders surveyed by Deloitte to invest in greater AI capabilities
57
in the next two years.
56
Growth of Business 43
78 • Use of Cognitive Technologies Such as NLP, Speech Recognition, Process Automation
and Robotics
34
Competitive Pressure 29
24
2 in 3 have planned investments in process automation such as
33  robotic process automation,
Increased startegic importance 35
49  customer resource management
2019  workflow optimization
31 
Expansion in scope of Service/
36 "next best action.“
Products 59 2017

13 2015
Expansion of Footprints 16
33

Survey results conducted by Deloitte in May,19

Survey results conducted by Deloitte in May,21 ///01
/3
Competitor Analysis
Com pany/ Aspect Dashboar Flexible
Product Product Offering Speed Scale Accuracy Based Multi- Social Multi- Entity d& Custom - deploym e
Nam e Analysis lingual Media m edia Extraction Reporting ization nt
•Language detection
• Linked entites
Microsoft •Language 50 requests/sec To be
Azure
•Named entity detection and
• Text -100 requests/second tested        
recognition
•Sentiment Analysis
•Natural Language
Understanding To be
IBM Watson
•Language Translator
Not Available
tested        
•Speech to Text
•Speech Analytics
To be
Marsview •Tone Analyzer Not Available        
tested
•Chat Analytics
•Call center operations
To be
Ameyo •IVR Systems Not Available        
tested
•Sentiment Analysis
•Call center solution
To be
Ozonetel •Sales Dialer Solution, Not Available        
tested
Business Phone System,
•Sentiment Analysis Social
Media Listening, Video To be
Repustate Not Availale        
Analysis, tes ted
•Named Entity Recognition

///01
/4
Solution Architecture

///01
/5
ML Model training for Mixed Language

*Image credit - Azure website

///01
/6
 Management Dashboard
Sentiment Snapshot Trend Analysis Region Wise Trend Analysis on
Sentimen % of total % of total negative sentiment
t calls calls last 1
(today) month
Positive
Positive 10 9 Some-
what
Somewhat 20 19 Positive
Positive Neutral
Some Some
Neutral 20 19 what what
Nega- Nega-
tive tive
Somewhat 20 22 Nega-
Negative tive
Nega-
tive
Negative 30 31
Day1 Day2 Day3 Day30 North South Central East West

Product Wise Trend Analysis on Performance monitoring of new Bank Wide Customer Positivity Index
negative sentiment Product Launches
Areas of
Concern

80
Some Some
out of
100
what what
Nega- Nega-
tive tive

Nega- Nega-
tive tive
Deposit Cards Personal Vehicle Home Day1 Day2 Day3 Day4 Day30
Loan Loan loan

///01
/7
 Customer Dashboard
Customer Positivity Index Trend Analysis of Emotions Movement of Negative Sentiment

• Desirable outcome on
the in the form of

65 downward trend in
negative sentiment
•
out of
Measurement of
Some- effectiveness of
what
Angry Annoyed Frustrated Upset Confused Hopeful measures taken by
100
Negative
bank

Negative
Day1 Day2 Day3 Day4 Day30

Product Category where Services were

not satisfactory Areas of
Intent Analysis Action Points
Concern
• Immediate feedback as well as
Switch Probability trend analysis
Credit Card Reduce existing products with Bank
• Well Defined actionable for low
Personal Loan
positivity index customer
Home Loan
No Change • Pointed Solution
Vehicle Loan
• Quick resolution
Increase share of Existing Product
Deposit & Other
• Prevention of dissatisfied
Day1 Day2 Day3 Day4 Day30 Scheme New Product customer expressing his
concerns on social media
Desirable • Cross sales
• Identification of
product/processes which is
causing pain to customers

This dashboards are only illustrative example. Many more dash boards can be created to get more granular details about customer, agents,
products and services etc.
///01
/8
Azure tools or resources

Azure tools or resources which are likely to be used by you for the prototype, if your idea gets selected

///01
/9
Key Differentiators & Adoption Plan

How is your solution better than alternatives and how do you plan to build adoption?
• The solution provided by W Technnovators will use AI and ML to provide real time assistance to the
agent about the engagement level of the customer. It will also use knowledge mining based on the
context of the conversation to provide resources.
• The system will also sense if the conversation is heading in right direction or an escalation to
sernior should be hinted.
• On supervisor dashboard a satisfaction widget will be available to show real-time status for the
ongoing conversations and quick access to historical data. Will also provide what level of coaching
assistance is required for each agent in different areas.
• On management dashboard organization level data will be provided based on access and
authorization.

///01
10
GitHub Repository Link & supporting diagrams, screenshots, if any

How far it can go?

///01
11
Thank You
Team member names
 References
https://www.mckinsey.com/business-functions/people-and-organizational-perform
ance/our-insights/the-moment-of-truth-in-customer-service

https://www.deloittedigital.com/content/dam/deloittedigital/us/documents/blog/blog
-20190513-2019%20globalcontactcentersurvey.pdf

///01
13