Welcome to an exciting journey through data.

In this video, we will unveil

stunning insights from our banking EDA project showcased in a beautifully crafted
PowerBI dashboard. Hey guys, Satyijit Patnik here and welcome back to my channel.
In data analytics, what you need is Python programming language, SQL as a knowledge
on databases, statistics knowledge, Microsoft Excel and PowerBI or Tableau. And in
this particular end toend project that is almost like a 2 and 1/2 hours of video,
you will be

getting a blend of all of this. I have spoken about data analytics portfolio
projects that they do not exist. you have to create it because there is no such
concept called as end toend data analytics project. In reality when you jump in
into a company as a data analyst you would potentially be doing either insight
gathering or data collection or data preparation not a blend of everything or not a
blend of insight gathering with dashboarding. But when you are in your learning
process, you have to utilize of what you have learned

and try to create your portfolio projects. And this video is going to help you
in that because this is going to cover all of these topics starting from Python,
EDA, SQL, statistics and PowerBI. To get started with we have taken a banking
related data set. So this is a banking related EDA project uh end to-end project.
We have taken the data set. We have cleaned it, prepared it, made it ready for EDA.
Then we did EDA using Python where we got some insight gathering and all those
things by doing variate, univariate and

multiple other analysis. And then we onboarded the data into MySQL database.
And after that we connected MySQL database from PowerBI as a data source and on
PowerBI we took insights from what we did in the ADA process and then we created
our final dashboard. So this is how we created this end to end portfolio project
leveraging all of the skills. So enjoy this video in case you have any specific
requirement on any other topic or any other project idea or any other video. you
know my niche I work on data analytics, data science, AI

and generative AI. In case you have any requirement, let me know in the comment
section and I will come back with a video. Apart from that, in case you like this
video, like this project, rate this project out of five in the comment section that
will be really helpful for me. Apart from that, those who are complete beginners,
you want to get into the field of data analytics as a career path, you are confused
what to do, you will be able to find a course link because I have a data analytics
masters

program which is at 4Ele 9 and I provide a lot of support there. 24x7 chat
support, onetoone touch-up points, um mock interviews, resume preparation and many
more. I will leave my uh WhatsApp number in the description. Check it out and you
can reach out to me for any kind of help or any kind of guidance. See you in the
project and [Music] enjoy. Hi, welcome to this new Capstone project video. In this
video, we shall be focusing on a banking related use case. Let's try to understand
what the problem statement is. The problem

statement is developing a basic understanding of risk analytics in banking and

financial services and understand how data is used to minimize the risk of losing
money while lending to customers. Now if you go to banks or financial services like
there are some private banks, there are some public banks, there are some other
financial institutions who also provide loan, right? So they provide loan to
customers. Now out of the customers, there are different types of customers, right?
There are some high earning

customers, there are some medium earning customers and there are some low
earning customers and they each have their own characteristics based on which they
are given some loan amount. For example, for high paying customers maybe the loan
amount the sanctioned loan amount could be around 10 to 50 lakhs. Just giving an
example. For medium let's say it could be like two to five lakhs. For low cap maybe
around 50k to let's say one lakh or two lakh right and this allotted amount differs
from customer to customer.

Right? This is very simple scenario to understand. Now out of all of these
customers doesn't really matter if you fall in this category, this category or this
category. There are certain customers let's say 5% of these customers let's say 10%
of this category and let's say 5% of this category they do not pay back which is a
loss to the banking company right or the financial institution right so our task as
a data analyst is going to analyze this data so basically do the risk analysis okay
Our work is to do the risk

analysis. Now before getting into this project, let me also tell you that
especially in data analytics, there is no such end toend project, right? You learn
different concepts. You learn Python, you learn SQL, you learn PowerBI, you learn
Tableau, you learn predictive analytics and then you yourself try to use all of the
skills to make it look like a end to-end project so that your profile is ready for
showcasing your profile in interviews or in the in the job hunting process. Right?
But in reality when you jump in to a company as

a data analyst you do not work on end toend project. There is no such end toend
project. Either your company uh gives you more work on the data cleaning side on
the data preparation side or maybe on the BI side on dashboard building side.
Right? So there is no such concept of end to end. But in this video we shall be
taking an end toend example so that tomorrow when you join a company you should be
familiar with all of these things. Whenever you are given a task whichever task you
are given you

should be able to accomplish it. Right now what we are going to do is we

already have data. Okay. So this is what my data looks like. I have different
client information, age, location ID, when did they join the bank, banking
contract, nationality, occupation, fee structure, estimated income because we
cannot have real income of the customer, right? So, we have an estimated income,
superanuation, savings, how many credit cards do they have? There could also be
another column like um um what do you say when you use a

credit card uh there is a civil score right like especially for Indian students
there is a civil score I'm not sure about whether it is applicable for foreign
students or not uh I literally have a very low experience on credit cards because I
have never owned a credit card uh but in India there is something called a civil
score right so if you spend a lot using your credit card and pay your amount on
time, your civil score is higher. And if your civil score is higher, then you are
provided with more loan because you are treated

as a loyal customer because you have been paying all your bills on time, right?
Something like that. So that can also be fetched. So it depends on what all data
available in the company. When you are starting working as a data analyst, your
task is to understand the use case and also try to get into the database team. Try
to talk to your peers, talk to your managers and try to collect as much data as
possible. In reality, data collection is a more difficult task. Okay. So we have
all of these columns. Now if you see there are

the last few columns are numerical columns, right? The relationship is provided
here. Gender one basically means male and two means female. Similarly BR ID is
banking relationship ID. One is retail, two is institutional, three is private
bank, four is commercial bank. Okay. Similarly the IID means who is the advisor
investment advisor and there are multiple people here uh Victor Dean Jeremy Potter
Ernest Knight and so on right so there are two ways either you directly uh you know
converge this data let's say

you just change it here to retail to as institutional now of course you can use
if else blocks to do Similarly, you can also change it here to make it just one
sheet and remove rest three of them. Up to you. Or if you want to create four
different tables, it's up to you. Right. Right. So, it's up to you. Now, this data
set is already provided to you. That means most of your task is already done. But
in reality when you jump into a company as a data analyst data collection is going
to be the major

task. Okay. And that you will realize with time. But can you mock up that
process uh without joining a company? No. Data collection is something that you
cannot mock up. That means you cannot recreate that particular thing. It depends
from company to company. Right? depends on what the company is asking for. Based on
that only we need to collect the data. Let's say your company is asking for uh
churning, bank churning, how many customers are getting churned out of a particular
banking scheme. So there are

different set of columns that you need in that use case. So similarly for
different use case there are different ask. Now in this case we already have data
set. Now our task is we already have data right what we are going to do is we are
going to load this data into a MySQL database right now in this process. So you
have learned all these things right. Let me write down whatever things you have
learned so far. You have learned Python. You have learned statistics. You have
learned SQL. You have learned EDA. You have learned MS

XL. You have learned PowerBI and Tableau. Right? Now, when you're loading the
data to MySQL, your SQL part is ticked. Right? You are using SQL. So, you will be
creating some tables. Right? Now what you're going to do is you're going to use
some sort of connector to connect SQL with Python. Right? So your Python part is
also done. In your Python, you are going to do data cleaning. You are going to do
data preparation and once this is done, you are going to do the EDA. Right? So your
EDA is done. While

doing EDA, you also use statistics. And as you're dealing with an Excel file,
your Excel is also somehow done. There is no need of performing data analysis on
Excel because we are already doing EDA on Python, right? And eventually once your
EDA is done, then you can leverage one of these BI tools to create a dashboard.
Right? So now this is an end toend project, right? No complaints. So let's get
started. Right. The very first thing is to convert this Excel sheet into a CSV file
so that we can have one

notebook and we will upload that into our database. Let me open my MySQL
database. All of these files will be provided to you. Nothing to worry about. So I
have my local instance. I have my company's instance. I'm creating a local
instance. Okay. So let's say I'm going to create a database. Create database.
banking banking case and the moment I refresh I should be able to see the table uh
sorry see the database banking case is created right here I will go to tables data
import wizard and then I will just paste

it okay we already have this and this is the CSV file yes banking Next. Now
table. Let's say I'm creating a customer table. You can check all of this. Next.
Next. So the data will be imported to your SQL database. Right now this will take a
little bit of time because the size of the file is little bit larger. Right now
this is how you can do it. Okay. I'm not going to show you step by step. Right? SQL
is taught to you, Python is taught to you. I'm just going to guide you in this
project. So let this process goes on. Eventually
 your table will be created. Okay. Next. Next. Your table is created. Show
tables. Okay. Sorry, my bad. Uh I have to use this database first. Use banking use
case. Sorry, banking case and then show tables and then I will show you the table.
Right, my table is created. All of this information is created. Right, no problem.
Now what I need to do is I need to connect my Python notebook to this MySQL
database. Now I'm using Google Collab. So what I'm going to do is I'm going to
switch it back to my local instance

because from Google Collab you cannot connect to your local MySQL database.
Right? So you would need a Jupyter notebook or a spider or um VS code whatever ID
that you are using. Right? Now once your data once your Jupyter notebook is open
right here you can actually create a new um new notebook and then here you have to
use PI MySQL or any other MySQL connectors. Python MySQL connectors MySQL connector
Python also you can use. Let's try to use this. Okay. Um I think we have already
covered the Python to SQL connectivity,

right? Let me get back the exact uh file. Uh okay, never mind. Um let's try to
use this connection and then get a [Music] cursor MySQL connector connect um okay
let's let's try to use this um connection equal to my SQL QL connector host equals
to this port equals to this user equals to root password is root connection is done
and [Music] then one second connection is done okay so let's say query is equals to
my SQL equals to df equals to pd read sql and then I also have to read put pandas
as let me write

it down and then if it works I will explain you the code pd readsql Then I have
to pass the query. Then I have to pass the connection. Okay. Connection close. Oh,
great. We have we have the database. We have the connectivity. So what I can do is
uh df head sorry my bad df head now see what I've done I've just simply connected
my database to my python notebook right this is what we do in our day-to-day
activity as well in as in the field of data analytics as a data analyst we could be
working with excel

sheets we could be working with CSV files we could be working with a database
Right? And when you work with a database, that is where your SQL knowledge comes
into picture. So here what I'm doing is I am creating a connector. This is how the
connection is created. Host is this. Now this is nothing but your local host. Okay.
The local host IP simply is 127.0.0.1. The port that we are using is 3306. In case
you um want to see I can always show you uh edit connection you can see and local
host right port is

3306 username is root and password is root I cannot show you the password but I
know the password is root and then this is my query right what is this query this
query is nothing but select star from database table name and then I'm reading the
SQL query using this connection and done my data frame is ready. Now once your data
frame is ready then you can take the next step right you can always do your
describe your df.info info and go ahead and do your analysis. Right? Multiple types
of analysis. So you have to proceed with

the EDA part. Right? So I'll switch to the collab uh notebook because I will
like to continue directly on collab. So for this exercise I'm going to use a
physical file which is banking.csv. But I have already showed you right how to
connect with my SQL uh database and how to work here right. So if you are willing
to work on local machine you can always do this connect to the SQL database and
start working on it. In case you are working on Google Collab you have to use a
physical file right just for practicing.

So let me also show you. Okay. So let me upload the file. And here we go. This
is my sif file. And let's get started with the EDA exercise. Right. So we already
have data. We already have data into MySQL uh database. We have already shown you
how to connect. Right? Now we will jump into this part. The data collection uh
sorry the data cleaning part, the data preparation part and the EDA part. Okay. So
to get started let's try to import all of our libraries. So import pandas as PD
import mattplot.pipot

pipelot as plt import se as SMS. If somebody asks you what is your end goal, I
think everybody knows your end goal. Our end goal is to do the risk analysis and
also try to show the important KPIs to our managers to our stakeholders. Right now
the important KPIs can be shown on Python ADA notebook as well. Right? But it is
not presentable and that is the reason why we take help of BI tools to showcase our
insights or analysis and that is the exact reason why we are also going to use
PowerBI as part of this

project. Not now after we are done with the EDA part. Right. copy path and then
df equals to pd dot read csv df head. So for now I'm just quickly doing um all of
these activities the moment I will be sending you out this python notebook it will
be more structured with more comments. Okay so the very first thing that I'm going
to do is I'm going to print the shape of it. The shape of it is 3,000 records and
25 columns. Simple. These are the basic things that I usually do with all types of
data set. You can see everything is not null.

So there is no such issues. And the major things are these. These are my
integer values. They are also not null. So I don't have any missing values. The
next thing is generating descriptive statistics for the data frame. And how do we
do that? DF dot describe. DF.escribe shows you all the statistical information
about all the numerical columns in your data set. It shows you mean, standard
deviation, the minimum value, the first quartile, second quartile, third quartile,
and the maximum values. Right? Now, how do we

interpret this data? Let's say for example, I'm talking about age, right? What
we can interpret is age varies from 17 to 85. Fine. and 75% of the people are above
the age of 69 years. Okay, let's say amount of credit cards. Amount of credit cards
you can see. Oh, amount of credit cards basically have concrete numbers in their
minimum and maximum value. So, this can be discarded as a numerical column. Okay.
The distribution is not proper because amount of credit cards is just one or two,
right? Some people have

one, some people have two, some people have three. Similarly, if you go to the
last sections, these also makes really no sense, right? The distribution is not
proper. So, similarly, you have to go through all of these foreign currency.
foreign currency you can see the minimum value is 45 the maximum value is this so
these are the statistical information we'll try to do some more analysis so that we
get to know um about um each of the columns so let's move on okay um what I'll do
is um let's say I will

pick one of the columns estimate ated income. Let's try to uh use estimated
income. Um df.estimated. Okay. Um what is the minimum value? 15,990. Okay, let's
try to box them. Um, what I'll do is I'll try to give them some boxes like 0 to uh
20,000 and then 21 uh 20,000 to 50,000. Let's try to box them. Okay. So, basically
from numerical data, I want to convert them into categorical data. Let's try to do
that. Uh so let's say I'm going to create my bins zero um 100k 300k any infinity
values um labels is equals

to low mid Hi. Um, DF income band is equals [Music] to don't remember. One
second. I want to create an income band from estimated uh income. I think it's pd
pd.cut. I mean I have to use pd.cut, but I don't remember the exact. Yeah, I think
this one. Yes, it's okay. If you don't remember code, it's completely okay. So, I'm
showing you the raw version of myself, right? I could have written down but
obviously I have chippity so I can take help of chity right so pd.ut Start
estimated income. Uh let me just

compare. Yes, estimated income. Bins equals to bins. Labels equals to labels.
Um yeah, should be okay. And then once this is done, I'm also going to um so we
have this column. Fine. Let me just try to do the value counts. Okay. And then plot
So high income we have this much low income and fine. Um now what I'm going to do
is I'm concerned about these three columns because these are not numerical columns,
right? These are also categorical columns. Anything else? Um number of credit
cards, amount of credit

cards, sorry. Yeah. Amount of credit cards. Anything else? Anything else?
Anything else? Anything else? Nationality. Okay. Nationality. Fine. Occupation.
Okay. Free structure. Okay. Let's try to take all of these columns. Let's say I
will write it down. Examine the distribution of unique categories in categorical
columns. Cat calls categorical calls is equals to DF of BR ID DF of gender ID. Uh
tree structure. amount of credit cards we did, right? Yeah. Properties owned.
Properties owned. And then what else? Properties

owned. And then risk weighing weight and br ID gender ID. We also did not take
IID. And then lastly risk weighing. And then um remember we did income band also.
We can take this also. Um and then uh let's say for call in cateorical calls counts
for Yeah, these are my value counts. Okay. Um, if you also want to plot, you can
easily plot. Oh, basically you cannot uh it's only showing you the last one because
fine. We can actually check the uh what do you say? Churn ML project churn. Okay.
Yeah, this one also can be done.

For example, let's say Let's say with respect to gender ID anyways let it So
what we are going to see here is the different value counts for each of these
columns right each of these columns that we have defined right for BR ID the
relationship ID we can see more from number three and then 660 so what was number
three if you go back to the excel sheet number three was number three was private
bank right so more records are from private bank and then from retail bank and then
from um institutional and then commercial

right gender you can see gender is equally distributed right 1512,488 equally
distributed IIA ID also equally distributed almost no there are certain customer
certain advisor who have 170 and certain who have half of them fine amount of
credit cards. Most of the people have one credit cards you know almost percentage
wise uh 1922 divided by around 2900. So around 66% of the customers have one credit
card right and rest is two and three. Similarly, nationality. Nationality most of
them are Europeans,

then Asians, then Americans, then Australians, then Africans. Occupation. Now,
this is simply a univariate analysis. You can also consider this as a univariate
analysis. That means you're analyzing one column at a time. Right? Let me go back
to the Okay, so we are still not able to run more than 20 figures have been opened.
Let me try to run only on these Yeah, you can see we are able to execute this and
this will give us a better idea, right? See colorings and all. I'm not focusing on
the beauty part of it.

You can always change the colorings. Take this to Google. Uh take this to chat.
ask how to make it more beautiful with u you know red blue colors whatever you want
to do for me numbers are more important than colors right so here if you look at
the distribution of male and female with respect to different BR ID you can see
this right as gender is equally distributed it makes really no sense because you
can see all of them are equally distributed but if you want to remove the ch remove
the hue also you can

remove it It's up to you. If you remove the hue, that means this is purely a
univariate analysis, right? And this same piece of code can also be used for
biariate analysis, right? You can see here gender equally distributed. Look at
these numbers. Amount of credit cards. Yes, we have more um people taking one
credit cards than two and three and then nationality and then goes on right for all
the columns. Now this is univariate analysis. I will try to write it here. Right?
Exact same thing can be used for

biariate analysis. Here you just have to add the hue. So what is univariate
analysis? Univar analysis simply means you are analyzing one column at a time,
right? You have a data. Sorry, my bad. Let the connection restore. Yeah, you have a
data and you're analyzing column by column. So first column the distribution is
this. Second column the distribution is this. Right? Third column the distribution
is this and so on. Right? And in the second piece of code here what's happening is
you are adding hue

that means with gender. So what you're doing is with respect to gender if I
talk about this one let's say this is um some column let's say okay I I'll talk
about a simple uh example let's say amount of credit cards okay now amount of
credit cards looks like this right 1900 something for one credit card and then 700
something for two credit cards and 300 something for three credit cards Right? Any
insight? No insight. Right? U just the insight will be there are almost 70% people
taking just one credit card. Fine. Now let's

say you want to analyze the same data with respect to gender. Right? So for one
credit cards let's say equal distribution almost 1,000 for male and 900 for female.
What is the distribution? Almost 50/50, right? Almost. Similarly, for two credit
cards, let's say 500 for male but 200 for female. Now, this could be an insight.
People who have taken two credit cards, most of them are males, right? And the
distribution is almost uh 50 by 70 will be 500 by 700 into 100 71% males. Similarly
for three credit cards 290 are males and

just 10 are female. That means almost 95% males have taken three credit cards.
So here the insight could be women are highly likely to take just one credit card.
Right? males are the ones who are taking more than one credit cards when required.
I mean I'm just giving an example, right? So if you go back to the code and then
run this with respect to gender ID, let's quickly go through all of the graphs and
see whether we can identify a pattern or not. If you look at the first graph, let
me go to the

top. Equal distribution everywhere. Go back. Equal dist. Okay. Here. Uh this
makes no sense. Okay. Here also. Okay. So two two is female, right? So female has
marginally more two credit cards. But I think it's still not an insight because the
number is very close. Male is around 370 380 and female is around 400. So it's
still okay. More female Europeans have taken credit cards. Good insight. It's very
clumsy for the occas occupation part. So I'm not looking at it. Everything looks
blunt. Properties owned. female

have more properties uh like one properties uh but not again not a very drastic
this thing let's try to uh analyze with respect to a different column let's say
gender we have already analyzed uh let's say I'm going to analyze with respect to
nationality T. Okay. Now we have some color this thing. I think I can see some good
uh this thing. Um okay. European Europeans have marginally okay. Uh the ratio 310
600 almost double 170300. Oh yeah almost double. Yeah I think it's following a
pattern. orange uh red is uh almost double of

orange in all the graphs. See, you have to take these numbers and compare. As
simple as that, right? Let's say this is around 160 and this is around 190 uh 290,
right? Almost double for this one 200 and then this one is around 120 almost
double. So not much of insights here. here European. Yeah, not much of insights
500. Yeah, this is where the insight could be. Most of the American and Asian are
having equal amount of three number of credit card but yeah it's still not a huge
insight. Now this is

how we have to go through it. Okay. So let me move to the next piece of code.
The next piece of code will be creating some sort of histograms. Let's say I'm
going to create a hist plot of value counts for different occupation for call in
categorical columns. If call is equals to equals to occupation because occupation
was little bit clumsy, right? Then continue. plt dot figure size 8 comma 4 seaborn.
Oh, seaborn. I might not have I already have Zbond dot um haste plot and then PF of
column uh pltitle histogram

histogram of occupation account. X label X label could be occupation, right? Or

maybe hardcode it. Instead of hard coding, just write it down. call while label is
equals to count. Uh so cool right? So what's happening here is um there is no count
on this rate. Okay. So analyze all of these graphs. Moving on to numerical
analysis. Let me create a textual this thing. Let's say numerical columns is equals
to estimated income. I'm just taking two of these columns. uh let's say estimated
income and superanuation savings in in your

case you can actually do this for all of the numerical columns okay so let's
say after defining I'm going to do univariate analysis analysis and visualization
plt dot figure size equals to let's say 84 for I column in enumerate numerical
columns plt dots subplot [Music] Okay. Um, why is showing both of them at super
annotation? Oh, okay. The first one is estimated income and second one is
superanuation. But why there is an overlap? One [Music] second. Is it something to
do with the figure size? Cool.

Yeah, now this looks good. Now you can see the distribution of it, right?
Estimated income and you can see it's skewed, right? This is also skewed.
Similarly, you have to also analyze different other columns that you have like all
the numerical columns that you have. Credit card balance. Let's try to do that
also. Okay. Uh three right credit card balance. Let's run it for all the columns.
Oh my credit card balance and then bank loans, bank deposits, checking accounts,
saving accounts. What else?

foreign currency account. Um, and then business [Music] lending. My bad. My
bad. Um, right. Cool. Um you can see somehow all of them are skewed. All of them
are right skewed. Right? Similarly, the very last step in the EDA process is going
to be um I'll try to do um some sort of heat maps so that we can also see the
correlation between all the numerical attributes. Right? So numerical attributes
are already defined. I'm just redefining it. There is no need but I'm just
redefining it. And then um heat maps SMS heat

map. So let's say correlation relation matrix plus df of numerical calls dot
core. So the correlation is done and then I have to plot it. plt dot fig size uh
plt dot figure size let's say I want a square one 12 + 12 and then sns do heat map
I'm going to pass the correlation matrix um annotation I want annotation true and C
map blue crest. can use a crest one. Uh uh cm map is equals to crest. Anything
else? I think that's it. Formatting is also needed, right? Format is equals to 2f
and then plt.title

title heat map or just correlation matrix plt.show. Now I should be able to get
a heat map. Wow, beautiful. Right now you can see that I can see a lot of insights
but I want you to perform this activity and come up with more and more insights. So
one of the insights I can see right now is you can see banking deposits has a very
high correlation with checking accounts and saving accounts right there is a this
is the strongest positive correlation right so banking deposits correlation with
checking accounts and
 savings accounts also with respect to foreign currency for foreign currency
account also around 0.41 41. This is also a strong one, right? This is also strong.
Even this is also strong, right? This basically indicates that customers who
maintain large balances in one account type often hold most of their funds or
substantial amount of funds across other account types as well. Right now talking
about credit card balance. Credit card balance estimated income. Credit card
balance and estimated income is not highly

correlated or bank loans also not highly correlated which could suggest credit
usage patterns vary independently of income level or loan amounts. Right? See if
you want to find all of these insights. Right? You can actually take all of these
output to your um chart GPD and you can also ask ch to generate insights. Now for
example, hey I want I want to generate EDA insights for my use case. I did a heat
map and I got a very high correlation between now you have to tell them right else
it will not be able to generate you

very high correlation between bank deposits and sorry my bad is closed bank
deposits and let's say checking account I'm checking accounts and checking um and
savings account, right? What could that mean? A high correlation between banks,
bank deposits and both checking and can indicate several insights. Financial
behavior. Customers who maintain checking accounts are likely to have saving
accounts and vice versa. This suggests that individuals who engage more with
banking services tend to manage their financial. This is

beautiful, right? You can actually leverage chat GPT to get down your insights,
right? So this brings to an end to the EDI use case. There are some of the insights
that we have got. Maybe I will try to write down some of the insights and as part
of exercise what you can do is you can create more and more insights insights of
EDA. Uh my first insight could be the strongest positive correlation occur among
bank deposits. Um bank deposits with checking accounts, saving accounts and foreign
uh currency account.

indicating that customers who maintain high balances in one account type often
hold substantial amount or funds across other accounts as well. Right? Similarly,
you can add more and more insights. That brings to end of our ADA. As part of next
step, what I'm going to do is I will tell you what is my thought process when I
take the next step of converting a ADA insights into a beautiful dashboard. In this
video, we shall be focusing on the dashboard creation. Use case is understood. We
are going to

develop a basic understanding of risk analytics in banking and financial

services and understand how data is used to minimize the risk of losing money.
Right? We already have discussed multiple things in the previous video. Right? We
had the data set Excel sheet. We converged it into a CSV file. We loaded into MySQL
database. Right? We learned how to use a Py MySQL connection or a MySQL connector
connection to read the data from a MySQL database to Python. However, in Python, we
did the ADA using a physical file, right? But I

have shown you how the connector is working. So, we did the ADA, data cleaning
and ADA. So this is what we did in the previous video, right? Now we already have
insights, right? We we know how the data is telling us and what the data is telling
us. Our next step is to basically take these insights and start working on a
PowerBI dashboard. PowerBI or a Tableau dashboard. And what is the motive of it?
The motive is very simple. As a data analyst when you are given some sort of data
to analyze by your manager by your

customers first of all you need to understand what and then do why what exactly
means the ED right what is the data speaking what is the data telling why is it
happening and then all the insights and when it is presentable when you want to
present it to the higher management or to the customer you need to make it look
presentable right Python notebooks are not presentable and that's why we go for
business intelligence tools where we are able to present our data now for this
exercise how does my brain function to

create a dashboard now this is where now this is an art from your data how do
you get into the dashboarding part this is where the art of dashboarding comes into
picture right now I'm going to tell you few things few ideas that will help you in
your journey whenever I start working on a powerb dashboard right what I do is I
first of all check the data and try to think of certain number of pages that I want
in the dashboard once I define the number of pages then I go ahead and do some
sketch After my sketch is ready, then I

actually go to Canva or a tool like Canva where I create the background images
and then I get started with PowerBI, use that background image and then onboard all
the numbers and all the measures and columns and blah blah blah. That is how I get
started with it. Now looking at this data, what is this data talking about? This is
eventually loan data, right? And in this loan data there are two types of uh
deposits right? If you see there is a loan right banking loans and there is also a
banking deposits.

What I can do is I can actually divide my use case or divide my dashboard into
such a way that I will first create a homepage, right? You can call it as a
homepage or a summary page, right? And then you can have like one page for loan
analysis and another page for deposit analysis. And last page you can have a
summary page. And last uh one more page you can have a ask a question page. So you
can have five different pages. Now let's try to see how does my homepage looks
like. Now what I'm going to do is let's say

I'm going to start working on the sketch. See this video could be like a 1 hour
long video right but in reality when you are working on a dashboard it usually does
not take 1 hour it take several hours to come to a final sketch and then from the
sketch to powerb dashboard it is actually like x number of hours of efforts that
I'm trying to converge it into 1 hour or 1.5 hours right so here what I'm going to
do is this is how my homepage looks like. I'm going to do a kind of a image here
like a logo. I'm going to write down here

banking dashboard. Dashboard I'm going to give a button for loans, a button for
deposit, a button for summary. And here if you want you can have some cards like
total number of clients, total loan, uh let's say total deposit, uh total fees, um
uh how many What else do we have? Income, amount of credit cards, credit card
balance. Okay, let's say some some let's say some metrics and then some metrics.
And here if you want you can give like a male and female kind of slicer. It's
completely up to you.

And then if you want you can also give a slicer for like the date like this
year this month something something like some people do January February something
something something people do year like 2023 2024 2025 something like that so this
is completely up to you but this is what I am aiming for now there could be another
layout it depends from person to person right another out can be simply a logo and
three buttons. All of these numbers not needed. I can potentially show these
numbers in the internal pages. No

problem. That is also possible. Right? And then after sorry after doing the
sketch for the homepage then you have to go ahead to the next step and start doing
the sketch for the next steps next pages right and once all of the sketches are
ready then you have to get started with the designing part okay so let's say I'm
going to quickly do a uh second page sketch as well let me try to do it here so
that we see everything on one page page. Okay. Let's say we are going to use this.
Okay. Okay. Done. This is my first page.

Now, let's say I'm doing a second page. And here I'm going to do a third page.
Second page. What I can do is let me have a look at the Excel sheet. uh and then
only I can have some ideas on on how my um sheet should look like, how my page
should look like. Um okay. So, okay. I can potentially show like total loans here.
Total loans. Uh I can show bank loans, business lending and let's say credit card.
And then here I can potentially show some sort of a bar graph like uh bank loan.
Let's say I can show bank

loan with respect to BR ID, business relationship ID and then here let's say
I'm going to show a pie chart uh let's say bank loan bank loan by income ID uh
sorry income range and then here I can show another graph let's say let's say the
graph could be anything Okay, something like this. And here actually I can give a
place for slicers. Let's say some slicer like gender, some slicer like uh banking
relationship and um let's say I don't know investment advisor maybe. Okay, that can
be done. And if

there is a space, what I can do is I can have a home button here and here I can
have a page navigation. So if somebody clicks on this they should be redirected to
this page. And similarly um the current page should look here and then the third
page should be here deposit. This also should be a page navigation and then the
summary page something like this. This can also be done. If there is a space here
you can have the date like here you are having year right? similar to that 2023,
2024, 2025, something like that. Okay. Now,

after this is done, um I wish I have more space. If I have more space, I will
like to add one more or two more graphs here. But let's let's do on the- go and see
how it looks like on Canva. Right? The third page could be like a replica of the
second page because it's nothing but the deposit analysis. So same uh page
navigation, page navigation and then here it's going to be a page navigation page
navigation everything remains the same. So here could be total deposit. Here could
be banking deposit. Here could be uh what else?

[Music] Um savings account. Savings account amount. And here could be checking
accounts. Okay. And here could be some graph. Here could be some graph. Here could
be some graph. And then finally the summary page. Summary page can also be
something like this. Um, summary page can have multiple cards like total clients,
total loan, total blah blah blah, total this, total that, total this, total that,
total this. So, let's say 12 cards and then some main filters here and on the top
remains the same. So, home loan deposit summary

and then like year something like this. So, now if I show you the graphs, let
me just take all of these towards this side. Okay. This is how my this is how my
four pages looks like. Understood? No doubt. And usually what I tell is when
somebody is in the initial stage of their data analytics career, they are getting
started to work on dashboards. I will always recommend you to Google uh help take
help from Google. Let's say you're doing a banking dashboard, right? Banking
dashboard. You all if you

just search there will be hundreds and hundreds of uh templates. Use any of
these templates. Whichever templates you like. Try to copy the idea. Doesn't really
I mean it's okay to copy the ideas. No problem. Let's say this idea is good. This
colors this color styling is good. This color styling is good. Up to you. Whichever
style you want to take, you can take it. Sometimes when you are working on
dashboards for a particular company. For example, I'm working for a company called
as Zep Analytics for example. Now
 you can see Zep analytics is mostly using uh yellowish and then bluish color
right. So potentially I will be using more and more blue color like this color
right this color you can see this color is mostly same right so I will be using
this color so always because eventually you are creating a dashboard for your
client not for yourself. So you need not have to be satisfied with the design, it's
the client who needs to be satisfied. So always think from the client's point of
view and if you're actually working on a

project tomorrow if you join a company, you are given a dashboard work try to
design some layouts, show them, get their approval and then work on the dashboard.
It should not be other way around. You design for yourself, you select and then you
build the dashboard and then show it to the customer. No, it should not be that
way. Right, moving on to Canva. Let's go ahead and we'll try to create our first
dashboard. Okay, so I'm going to create the first dashboard. Let's say I'm create
clicking on create

design. And here you can just simply do thumbnail YouTube thumbnail done. And
then here you get started with your design. So let's say the very first thing is
let's say uh I want to take let's say a banking dashboard logo. Okay, let's say I'm
going to take this image. Okay, now I have taken this image, right? And then I'm
going to create some boxes. Let's say color I want to give like dark colors. See
the more time you spend in this particular design the better design it will be
right as simple as that there is

uh no doubt about it right so you have to spend more time to it in my case I

think I need to upgrade I might not have uh Google uh sorry might not have Canva
pro I used to have but somehow I think it's gone I don't know. I have to check it
out. Take up here. Okay. And let's say my boxes are something like this. Okay. H
and then um Okay. And here I can write like banking dashboard. Banking dashboard.
Let's say I will try to make it like 30 or something. Looks fine here. If you want
capital you can try

this. Cool. And then loan deposit. And here I need to use white. And I need to
make it something like that. Deposit analysis. Then make it little bit big. Make it
little bit big. Okay. And then make it white dark. summary done. Doesn't look
really bad. I think it's fine. And then on top of this, maybe I can try to give
another box where I can pass here something like that. And here let's say gender 15
total clients uh let's say at the middle of it and then something here then
Something here. Then something here. Let's click here.

Then something here. Um, what else can we do here? Total clients. Yes. Fine.
And then total nationalities, I think. No, not needed. Total loan can be done. We
have a loan amount. Bank's loan. Total deposit can also be done. Um total fees can
also be done, total CC amount, savings account amount, checking amount. Yeah, let's
let's just write down total loan total deposits. After total loan and deposit,
let's say checking accounts, savings accounts, and then let's say business lending.
Okay, something like that.

Right now I have to take Okay, it looks I should make it something like this.
Looks fine to me. If you want to give a d background color, it's up to you. But I
don't think there is a need of background color. Maybe I'll just keep little bit
like grayish. I think this looks fine. Yeah, I think looks fine. And banking
dashboard page one, right? And then I will click on save and download. And then
I'll go to PowerBI, open the PowerBI desktop. My first background image is done.
Similarly, I can work on multiple other

pages. But let's try to focus on the first page. Okay. So, going back to
mirrorboard, I hope everybody remembers that the data is already on MySQL, right?
So, when you are actually using PowerBI dashboard, you can either directly use a
MySQL connection or directly use an Excel file. The choice is yours depending on
what your company is asking for. To practice a end toend portfolio project, I will
recommend you to use MySQL database so that you have a practice with a MySQL
database. But for now, it's completely up to you. Okay,

let me quickly start with a blank report and show you how to connect with the
database and then I will actually eventually I will going I I will be working with
a Excel file. Okay. But I'm just showing you how to get connected with the MySQL
database. Okay. So let's say MySQL database. I click on this. My database is
active. Fine. Server is going to be local host. Database is going to be my banking
case. Advance not required. I will just select star from this. That's it. Uh
banking case. Okay. Now you can see I'm able to get the

client uh sorry the table and this is it. My table will be loaded and then I
can eventually start working on it. So let's say my first thing is I will be going
to format and then canva background. Use my dashboard page one as my first page. In
transparency I will make it this and then I will make it fit. That's it. Once this
is done you can eventually pass your customers. So where is the customers? Let's
say I'll go back to this table. Name right? Where is name? Wait. So here I will be
adding a

card and then I will be adding name and then I will be making count 3,000
customer names. And here if you want you can always go to this formatting options
and then go to this general and then change all of these uh effects. Let's say you
don't want the background color you want to uh change the title turn on turn off
effects header icons these kind of things. And visual let's say you don't want the
category label. So something like this and then here goes the number. That's it.
Okay. Total number of clients. And

this is how things should start looking like this. Similarly, you can always do
total loan, total deposit, checking accounts and so on. Right? Gender. Let's say I
want to give a filter on gender. So, gender I'm going to do. Okay. One thing that I
can do is so right now my gender is 1 2 something like this right either you apply
here you just go and replace values one with male and so you cannot replace because
it's a numerical attribute first change it to a textual part and then once the
textual

part is done then change it to male and then change to to female. Done. Either
you can do this or you already have that mapping table right the gender ID one and
male two and female you can take that also and then using that also using if else
block also you can do the calculated column it's up to you so I'm not going to get
into this these are very basic basic things of power P and I expecting you to
already know about these things right going to visualization I'm going to use a
slice slicer here. So the

slicer is done. My gender ID comes here. And now here I'm going to use a um
slicer. There was a slicer style tile. Done. And then I'm going to remove this
header icon. Visually slicer header is gone. That's it. Click on female 1512. Click
on male 1488. Click on none of them original. Right? Beautiful. Right? So you work
with a database or you directly work with an Excel file. It's completely going to
be your call. Right? So let's say you you want to work on your own this thing you
can always select which let's say you want to check

this one and table two table three table four relationship clients okay and
then then up to you so right now I'm just using a SQL database And I will be using
the SQL database for rest of the things. Right? So my homepage is done and then I
will move to the next three pages. So let's move to the next three pages in the
next video. Before moving into the next part, actually I forgot to um explain one
more thing that going back to um our design, right? So we have total clients, we
have total loan, we have total deposit,

um checking account, savings account, business lending. Okay. So okay first of

all total loan right um how do we get total loan because if you look at this entire
sheet right there is a column called as bank loans but total loan is not the
meaning of bank loans right so what I'm going to do is I'm going to take help of
Chad GPT because here comes your domain knowledge right can Can you tell me the
formula of total loan and total deposit? Total loan formula is okay. It is showing
you the summation of bank

loans, summation of bank deposits. Okay, fine. People who do not have that kind
of knowledge on banking, it's completely okay. As per your knowledge, you can
quickly go ahead and do the banking loan um to this this one. It's completely fine.
No problem. So, let's say I'm just trying to do the same thing. Uh sum of banking
loan is going to be 1.77 billion which is fine. No problem. But in case you know
the actual formula is different. The actual formula is let me just try to write
down uh total loan is nothing but bank loan.

Bank loan is sum of bank loan plus business lend lending business lending plus
sum of credit card balance. Now this is actually the formula of total loan. Okay.
Now here wherever I see the loan where is that total loan I'm going to change it
here instead of this I'm going to use this and this becomes 4.38 billion if you
think I'm not correct I you can potentially go ahead and ask charge GPD okay the
formula is correct and and I know this formula because I have worked on banking
sector so I know this and then it's completely

okay if you don't know this formula because this is a very very domain specific
formula. Okay. Similar to this for total deposit also. Total deposit is summation
of bank deposit plus sum of um savings account plus sum of um bank deposit savings
account. and deposit savings account, foreign currency account and checking
account, checking account. So the bank deposit, savings account, bank deposit,
saving account, checking account and foreign currency account and then this is your
total deposit and then I will bring on my

deposit total deposit 3.77 billion. And similarly last three of the fields you
can fill it up no problem. Okay. Um, savings account is nothing but total of
savings account, total of business lending, total of checking accounts and so on.
And here you already have this thing which is also fine. Do we have like a column
for date? Joined bank. Joined bank is there. and uh join bank. These are the ears,
right? Maybe maybe I can show some years here. Uh let's say I'm going to add a
slicer. Okay, let's see. I'll first

create a column. Okay. Column is year of joining is equals to year of year of

joined bank. And then I will try to create a slicer like this. And then I will make
it short. Make it like this. And then I will add year of joining H. But here I want
to show only the first five. It's okay. I can show this. No problem. So let's say
2021 the numbers are this 2020 the numbers are this 20. So beautiful right? So one
thing that I'm missing here is uh nobody will be able to understand what this ear
is right. So what I'll do is I'll try to

make it look like this. And here I will add the ear. So I'll again go back to
my um this thing like my Canva and add here joining here. Right. Right. Joining
here looks fine. And then again download it. And again go to this background Canva
background. Change it. Download the latest one. Done. Transparency zero. Fit.
That's it. Looks good. And then here I will just make it background off and remove
it like this. Um here actually I can show the top values. Um one second I'm just
forgetting how to do

that top. I think as a filter I can add filter and I can only have Um top end
sorry yes top end top five and then by value by year of joining apply filter and
then I'm able to see the top five yes right great now I'll hide this filters and
visualization and data and this is my my first page is ready okay why do we keep
the fields empty Let's just finish everything together. Checking accounts. Checking
accounts. So I will be adding instead of deposit checking accounts. Instead of
savings account, savings

account, instead of business lending, business lending done. Got it? And then
my first page is completely done. Similarly, I'm just naming it. Let's say banking
dashboard 25. Save it. Then let's move on to the next pages. Hi. Moving on to the
next part which is loan deposit and uh deposit analysis pages. Sorry, loan deposit
or loan analysis. My bad. It was loan analysis. Loan analysis and deposit analysis.
One more thing before we jump into the next part is I kind of like this. I would
also like to try one more

type of um background image. So, let me just copy everything and uh make the
background image. How do we change the background image? Come on. Uh, wow. Wow.
Let's say blue color. I Yeah, I want to make it something like this. And then on
top of this, let's say I want to have I'm not sure. I'm just giving it a try. As a
designer, you should give multiple tries. And then okay then I can actually change
this color to this color. Wow. Good. Copy this and this one to this color. This one
to this color. Looks good. And this one. Let it

be. and then Okay. Okay, not bad. And then I'll move this one as well. And move
this one as [Music] well. Okay. Okay, now this looks like a new homepage. Let me
get it to the PowerBI dashboard and see how it looks like. So, let me just go here.
Go to visualization page canvas. Change it to the new format and make it fit. And
shift all these a little bit so that they are in sync. And then this one something
like this. And this one will be something like this. Now I'm not sure whether this
one is good or this

one is good. So I have done two different versions. Maybe you can decide which
one is better. Let's say home one and home two. Okay, for me home two is better. So
I'm going to use this as my baseline and I'm going to create another page. Okay.
So, I'm going to use the same uh styling and then here, sorry, I'm going to okay,
let me design and then I'll try to explain you um where the things are and what the
fields are. Right? I'm planning to do something like this. Right? I need to add
these buttons as

well so that they are visible. So I'll try to add some text here. Actually we
need to copy this style. Sorry, this style. And of course I'll make it little bit
smaller. So I have my homepage. I have my uh which one is that? loan analysis. I
have my deposit analysis. Cool. And then finally, I have my summary. Doesn't look
bad, honestly speaking. And then I can have this joining ear uh somewhere here. And
I'll make it like this. I need not have to write joining here because of the fact
that it's already mentioned in the homepage. So

it's already expected somebody who's going through a dashboard. He already

knows that this is this year is nothing but the joining year, right? So I'm not not
going to add it. If it does not look good, I'll try to add it later. Okay? For now,
I'll keep it this. Now here in this box I'm going to add uh some um slicers. So
I'll keep it empty for now and I will take this image to PowerBI and then I will
see how it looks like and again come back here. Okay. But this box looks fine to
me. And here I

will be adding one box for total loan, one box for um something and then let's
say I will have let's say uh I'll keep it like this and [Music] And not bad. And
then uh this side I will have one more graph H. And then let's say I will have one
graph here, one graph here and let's say I will have one graph here. This is my
layout. I'm planning it this way. And then I'm going to add total loan. So this is
your loan analysis, right? So loan analysis. Now this has to be um this has to be
on top of like a

rectangular this thing. You're getting my point, right? It it should look

different. Cool. Looks good. And then I will add all these names. Let's say total
name. Total loan. Remember what was my total loan? My total loan was um my total
loan was uh bank loan, business lending. So let's say this is your total loan. This
is your bank loan. And what else? We have business lending. Business lending.
Lending. Then we have credit card balance. Right? So I will also show the credit
card balance here. CC

balance. Okay. Done. These are my four attributes. And here let's say I'm going
to show some graph. See, honestly, I don't know which graph I want to show here. So
for now I'm just writing it as uh bank loan by income band. Make it left aligned so
that we can align it like this. Done. And similarly here um bank loan by bank loan
by bank loan by okay by banking relation bank loan by BR okay and similar to this
let's say total loan by BR Or what else can we do? Bank loan by VR, bank loan by
nationality, by occupation.

Let's do nationality. By nationality, and this one, let's do by occupation.

Okay, now this looks good. I'm going to download it. I'm going to download only
page number two. Done. Download. And then I'll go here. I'll create a plus. And
then I'm going to use background browse. And I'm going to use the new image is not
yet downloaded. Oh, it's already downloaded. And then transparency fit. Done. Try
to Okay, looks pretty fine. It's asking to sync visuals. Yeah, let me sync it.
Cool. All good. Now firstly let me name it as loan

analysis. Secondly I will be adding a card here. Right? Basically I will be

inserting a button here. Let's say I'm going to insert a blank button here. And
this button should not have any title, any background, any border because it should
not be visible. But I can see there is a border. So I'm trying to find where this
uh styling is. Oh, icon border everything I turn off. Everything is gone. Now here
I want to go to action on and then type is page navigation where I want to navigate
to loan analysis. Now just to test it out

control-click we are going to loan analysis. Right here also we need to add the
same thing and here also I will be turning off the styling and same I will be
adding on deposit analysis same I will be adding on summary okay so here I'm going
to use action page navigation where home to done Deposit analysis. Let's say I'm
going to create a new page. Deposit analysis. Similarly, I'm going to create a new
page which is summary. And let everything be empty right now. Deposit analysis
action one page navigation. Deposit analysis

summary action. Page navigation summary. Right. All good. Let's test it out. If
I'm on this page, I'm I can go to homepage. I can go to loan. I can go to deposit.
Fine. And summary. Uh let's say from here to summary. Yes, we can go home to to
deposit. Uh so this part is pending. So I will just use this action page
navigation. I will make it deposit. Same I will copy paste for summary. Action.
Deposit. Summary. Okay. Test it out. Summary. Yes. is able to go dep uh sorry
deposit. One second. Oh, did I not copy paste

it? Oh, I did not copy paste it. My bad. Okay. So I again need to create blank
buttons for loan style turn off turn off point and then action turn on page
navigation loan analysis and same again for deposit analysis style turn off turn
off text style action Page navigation, deposit analysis. Test it out. Loan
analysis. Go back home. Deposit analysis. Yes, working. Summary. Yes, working.
Right. Everything is working. Now, going back to loan analysis. What else do we
need? We need to add total loan here.
 So, I will be adding a card here. Actually, I'm going to copy paste the card
from here because the formatting is already done, right? So I'm just going to copy
paste it here and then um all good. Here instead of this I will just write down
total loan. Okay done. Similar to this I will be using bank loan. Bank loan.
Similar to this I will be using business lending. Business lending similar to this
I will be using CC balance. credit card balance done. The combination of this
should be this 1.77 billion 2.6 billion

is around 4 37 billion and this is like 9 something billion. So 4.38 billion

right looks fine. Now let's say here bank loan by income band. Now income band do
we have an income band? If you remember uh we had actually created an income band
in our EDA right similar to that we need to create an income band right now. Okay.
So what we can do is um go to my transform data income band estimated income. Okay.
So let's try to add a conditional column. Let me go back to Google Collab. I don't
remember what was my logic. I

will be using the same logic for this uh PowerBI dashboard as well income band
okay and the logic is okay 100k 300k something like that so I'm going to use if
estimated income is greater than or equals to 300k output is going to be high. else
if estimated income is greater than or equals to 100k then it's going to be med uh
actually I don't need this else it's going to be low and then click on okay and the
name is going to be income band done income band and then go back close and apply
and then go

back to my PowerBI dashboard and here I'll be creating a pie chart or my baby a
donut chart and in this donut chart I'm going to add why it's custom I I I renamed
it to uh income band Right. Income band. Yeah. Why? Maybe my mistake. So let's say
here I'm going to use income band. Income band is going to be the legend and uh
data is going to be bank loan. So loan is going to be the values. Okay. And this is
how it is. I'm going to turn off the title. Sorry, not legend, the title. I don't
need the

title. I also don't need the background. Uh apart from that, I also don't need
to have um maybe I'll just try to change the values a little bit so that the values
are like this. And uh fine. And yeah, this looks good. Similar to this, I can have
bank loan uh by BR. That could be like um instead of this maybe I can have a BR ID
as a legend. I don't need the legend. I want something like this BR id uh BR id
looks like 1 2 3 4. If you want to change it to the concrete values, you can
actually use them like

use if else block to either convert this column to this or else you can also
use replace values. Right? For example, if you go to transform data, there are
multiple ways to solve a problem, right? You can choose which way you want to
solve. Let's say you changing it to text and then you can replace values one to be
retail and then two to be institutional two to be institutional. Private bank is
three and then commercial is four. Up to you. and then close and apply and go back
to your PowerBI and your data should be updated.

Good. Okay, something like this. And similarly, you can also have it bank loan
with respect to nationality. And here if you want to show like couple of
nationality you can you can show like the top nationalities or something like that.
Let's say you want to show only the top three records based on the loan values you
can always show that up to you right European and Asian and American. And lastly
there is occupation. Here also you can have some sort of graphs or maybe whatever
you want to show. Right? And then similar to

that um one more thing is I need to have slicers. So I'll create my first
slicer where the first slicer should be uh be and I want to change it to let's say
tile. Right? So something like that. And then I'll shorten it up. This looks like
this, right? Sorry. The banking relationship. See, it's up to you if you want to
write this name or you want to delete it from here by removing the slicer header
and going to Canva to make sure that the form the the the font style is same. It's
up to you. It's your call. Let's say I will

keep it like this. And similar to this, let's say I want to have something for
gender. I will be adding gender, male and female. And lastly, investment advisor. I
think I'll keep it this way. and investment ID I want to make it drop down okay
something like this and again instead of 1 2 3 4 you can always replace the values
let's say I want to change the slicer header it looks like this I bring it down I
bring it little bit down bring it down bring it down okay so I have space for
something. So I'll go back. I

will try to add something here. So let's say I want to make it um uh banking
relationship, gender and Institution advisor. Select pages. Just second page.
Download. And then go back here. Go to canvas background. Change it. Browse. Add
it. Fit. It. That's it. Done. Looks good. Wow. I I think it looks good. Right now
what I need to do is in similar fashion I can always do a duplicate and create a
Right. Similar to all these details, we can have total deposit um bank deposit,
saving account amount and here we can have checking

account amount. Okay. amount amount. Okay. Similarly, bank loan or sorry, it
should be bank deposit by income band, bank deposit by VR, deposit by nationality.
Similarly, bank deposit by occupation. Okay, now this is done quickly. Let me
download it. I will also create um let me download it first page. H go back to this
one. I will go back to I will actually try to create duplicated So I will just move
it this side and here I will change my color this thing to this thing. That's it.
Done. Um you

know what to change right? So here instead of all of this total loan and all
just change it to deposit total deposit right? Uh, banking deposit, change it to
bank deposit. Savings account, changing to savings account. Checking account,
change it to checking account. This one, bank loan. Bank deposit. And similarly
here. Similarly here. Similarly here. Right. Cool. Um whatever graph I just giving
you some ideas, right? Okay. Right. Deposit analysis is done. And here I will be
adding these things. Homepage. Yeah. These things are

done. Deposit analysis. Okay. Here I need to change the action to deposit
analysis. Here I also need to change this to deposit analysis. Deposit analysis. I
need to change I need to have a button here for loan analysis. done. All good. So,
my homepage is done. My loan analysis is done. My deposit analysis is done.
Summary. I'm going to do the quick um um uh graphical representation of it. and
then I will try to open the live dashboard that we have already created. So we
already have created a PowerBI dashboard just as an

exercise how to get started on creating a dashboard. I was creating from

scratch and showing you all of these things. Right? So summary page I'm not going
to create it's going to take a lot of time. So I'll just quickly show you how to
create it. Right? So let me delete all of these things. My bad. Um, so let's say
I'm going to just add basically I need to make it a little bit bigger. Okay, this
is being a little bit bigger. So, I'll just decrease it. Not bad. I think now we
are able to do cool. I think this is

good. This is going to be my summary page. Right? And all of these boxes I will
be writing whatever important metrics we have like total loans, total deposit,
total blah blah blah. Right? There are so many numerical attributes to show. So I'm
going to show everything here. And as part of a last page, what you can do is you
can actually do a ask a question page. Right? So let's say how do we create a ask a
question page. Simply doubleclick, right? And then your page is ready. So this is
how my summary page

looks like. just to show you guys actually I have done most of the things so
why don't I just uh total total clients total loan and blah blah blah. Okay, one
more thing actually al you can also add icons here. Okay, let me also show you
that. See, I'm I'm forgetting a lot of things because there are actually a lot of
things. You can actually make it make it even more beautiful by adding all of these
things. So, let's say I'm going to add like icons here, right? If you show the
numbers and then the icon here looks good, right?

Similarly, let's say loan total loan, then you can actually show something like
this. Right. Total deposit uh deposit deposit deposit. Oh, there is actually icon
for everything. Similarly, accounts. I'm just showing it for these four. I think
you get an ide you got an idea right this is how things work maybe I have done
everything why should I just leave two pages then so let's say I have my savings
account so it looks like this and then my business lending it looks like this
beautiful of course you can make it

black and white or color whatever you want to make it um most of the things are
done and the homepage. Anything else? [Music] Um yeah, also if you want to add
these things Now, this looks good. I'm just downloading this first page. Um, okay.
Let's say just for the sake of showing you guys. I'm just downloading a watermark
version. Of course, you would be needing a pro account to um get things sorted. But
I think that thing you should figure it out, right? Not going to interfere a lot
here. Just restructure it so that it

looks better. Cool. Now our dashboard is ready. Now our exercise is over. I'm
going to show you the actual dashboard that we have created. So let me switch back
to the actual dashboard. In the actual dashboard actually we are using Excel sheet
as one of our sources and not directly from MySQL database but I think you will
figure it out right things are pretty very straightforward as well and one more
thing you can see here is uh there is a if you remember in PowerBI we discussed
about a best practices right so here they have

actually done the best practices all of the measures are actually created
separately in this particular uh new query, right? Which is good. But in in our
case, we have done everything inside this particular query which is also fine
because here we just have one query but here there are multiple queries, right?
They have taken all of these gender, investment, banker, advisor. So they have
taken all of these um uh sheets, right? But you can see this is how the graph this
is how the dashboard looks

like. As I told you creating the first version of your PowerB dashboard could
be like one or two hours. But if you want to make really a beautiful beautiful
dashboard, you have to spend a lot of time in uh you know fonts in in the color
selection in restructuring. So there are a lot of things right in two one or two
hours it's not possible to work on end to end projects but I'm able to do that. So
that's all about this particular project. I hope you enjoyed this project and u
yeah that's it. The codes and everything will

be provided as part of resources. See you. [Music]