The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.

This is the Final Report of Industrial Training of WIE170701.

Discover the best professional documents and content resources in AnyFlip Document Base.
Search
Published by jasia t, 2020-01-31 00:58:54

Jasia's Inrernship Final Report

This is the Final Report of Industrial Training of WIE170701.

Keywords: Internship Report

FACULTY OF COMPUTER SCIENCE
AND INFORMATION TECHNOLOGY

UNIVERSITY OF MALAYA
I​ NDUSTRIAL TRAINING FINAL REPORT

SEMESTER 1​ ​ SESSION ​2019/2020
JASIA TUNAZZINA
WIE170701

SUPERVISOR FROM THE FACULTY:​ D​ R. HOO WAI LAM

1

i.​ ABSTRACT

This internship report serves the purpose to record the details of my industrial
training which was conducted at Freakout Malaysia Sdn Bhd. and Kaodim Sdn.bhd
accordingly. This report will cover the details of my internship in the system analyst
and data analyst department for a duration of six months which began from 26th
August,2019 till 21st October,2019 at FreakOut Malaysia Sdn Bhd and 29th
October,2019 to 14th February,2020 at Kaodim Sdn Bhd. During the period of
industrial training, I was supervised mainly by Mr. Kazuya Naraoka at FreakOut and
Mr. Yan Leong Woo at Kaodim.

In this internship report, I will elaborate about the experiences and
occurrences which I had encountered and the things that I learnt during my entire 24
weeks internship period which includes data analysis, data entry, data mining, data
visualization, uploading in websites, analyzing the systems etc. Besides that, an
overview of the company and the clients’ company which i had gone through will be
illustrated in the report.

My experience as a system analyst and a data analyst intern has benefited me
in many ways and one of the greatest benefits for me is to be able to gain valuable
experience by getting involved directly in the tasks related to data and system
analysis. Those first-hand knowledge and experience that I gained are not something
which i can simply learn from textbooks or lecture classes. Besides that, I am able to
apply those knowledge and theories which I had learnt in the classes to my daily
work.

In overall, this training program is a good learning opportunity for me. Now I
am able to expose myself to the corporate world and technical knowledge regarding
data and system analysis. I also have the chance to apply the theoretical knowledge
that I learnt in UM into the task given. Therefore, I found this industrial training
extremely helpful.

2

ii. ​ACKNOWLEDGMENT

I am feeling grateful to be able to go through my internship which lasted for 24
weeks with FreakOut Sdn Bhd and Kaodim Sdn Bhd accordingly. It was a
memorable experience for me as I got to learn a lot of things during the entire
program. This industrial training had helped to strengthen and develop my soft skills
such as communication skills, time-management skills and interpersonal skills.
Therefore, I feel gratitude to all the wonderful people who contributed and helped
me in different ways throughout this period.

First of all, I would like to thank our coordinator of the Industrial Training
Programme, Dr.Aznul Sabri and my supervisor, Dr. Hoo Wai Lam in assisting me
throughout this internship period and also greatly appreciated. Whenever I had any
doubts and any problem they will be there to answer my questions patiently and help
me to clarify those matters which I was unsure of.

Besides that, I would like to express my greatest gratitude to FreakOut Sdn
Bhd and Kaodim Sdn Bhd which had provided me the wonderful opportunity to
undergo my industrial training with the firms. I would like to thank Mr. Kazuya
Naraoka from FreakOut for being so helpful supervisor and his guidance throughout
the period. I would also like to thank Mr. Jamil Azri for hearing me out whenever I
faced any difficulties during my tasks and help me with the solution and to achieve
better understanding.

I am expressing my deepest gratitude to Mr.Yan Leong Woo who had
provided me supervision and guidance at Kaodim Sdn Bhd. In spite of being
occupied with his job, took time to hear, guide and keep me on track as well as
allowing me to carry out my project at their esteemed institution. He and all my other
data team had helped me to achieve a better understanding in the data science field
by sharing their knowledge with me. Moreover, their friendly behavior had made me
feeling warm and grateful to be able to work with them under the same engagement.

3

In short, I would like to dedicate my special thanks to my family and my
friends for their continuous support and understanding during this period. It will be
somehow stressful to work in a data science field and also it was stressful during this
period due to the cessation of my first company FreakOut due to their necessity to
meet the parent company’s KPI responsible for its shareholders and work loads and
timelines. Nevertheless, their encouragement helped me to face the challenges which
I met with more strengths, patients and positive mind.

4

​iii. T​ ABLE OF CONTENTS 2

i​. ABSTRACT 3

ii​. ACKNOWLEDGMENT 5

iii.​ TABLE OF CONTENTS 7

4. LIST OF FIGURES 9
9
5. OVERVIEW OF ORGANIZATION
FreakOut Malaysia Sdn. Bhd ( The Appster Project): 10
10
6. THE APPSTER WORK EXPERIENCE 11
6.1. OVERVIEW 12
6.2. TRAINING 17
6.3. DATA ANALYSIS 19
6.4. SYSTEM ANALYSIS
6.5. WEB DEVELOPMENT 20
20
7. SKILLS GAINED 21
7.1. PROBLEM-SOLVING SKILLS 22
7.2. ANALYTICAL SKILLS
7.3. SOFT SKILLS 23

8. OVERVIEW OF ORGANIZATION 23
23
​Kaodim Sdn Bhd. 23
8.1. BACKGROUND OF THE COMPANY 23
1.2 LOCATION 23
1.3 VISION 24
1.4 MISSION 24
1.7 BRANCHES 25
1.8 PARTNERSHIP
1.9 ORGANIZATION STRUCTURE 26
26
9. KAODIM WORK EXPERIENCE 26
9.1 SOFTWARE COVERAGE 26
9.1.1. Analytical Tools 27
9.1.2. Data Visualization Tools 28
9.2. TASKS RECORDS 29
9.2.1. Vendor Sourcing using Google Trends 30
9.2.2. CRM Segmentation using RFM Score 32
9.2.3. Country Analysis (The Philippines)
9.2.4. Category Expansion 5

9.2.5. Update data on Facebook Using Facebook Ads Manager 33
9.2.6. Link Updating using Branch 34
34
9.2.6.1. Facebook Product Catalog Link Update 35
9.2.6.2. Unbounce CTA link update 36
9.2.7. Connecting Google Map API and Deduplication using Python 37
9.2.8. Creating Decision Tree using SPSS 38
9.2.9. Campaign Strategy Planning 39
9.3. PROJECT INVOLVED 39
9.3.1. Grab-Kaodim Project
40
10. SKILLS GAINED (KAODIM) 40
10.1. PROBLEM SOLVING SKILLS 41
10.2. ANALYTICAL SKILLS 42
10.3. DATA PRESENTATION SKILLS 43
10.4. SOFT SKILLS 43
10.4.1. Communication skills 43
10.4.2. Critical thinking 43
10.4.3. Teamwork 44
10.4.4. Time Management 44
10.4.5. Strong Work Ethic 44
10.4.6 Self-Confidence
45
11. SUGGESTION
46
12. CONCLUSION
47
13. REFERENCES

6

4. L​ IST OF FIGURES

Title Page Number

Figure 1: The nature of the work. 10

Figure 2: The PUBG API Documentation 12

Figure 3: The buckets Created on Amazon S3 13

Figure 4: The SQL Query done on AWS Athena 14

Figure 5: Python coding using PUBG API on Google Colab 15

Figure 6: The flow of data to do automation 16

Figure 7: The system invalidation process Amazon CloudFront. 18

Figure 8: The website created using NEXT.Js 19

Figure 9: The organization structure of Kaodim 25

Figure 10: Use of Google Trends to analyze data 28

Figure 11: The RFM model for CRM Segmentation 29

Figure 12: The holistics report made for the Philippines 30

Figure 13: The hypothesis and findings about the Philippines 31

Figure 14: The Power Bi report created by analyzing data 32

Figure 15: The Google Sheet file for category expansion 33

7

Title Page Number

Figure 16: The use of Facebook Ads Manager for customer list 34
update

Figure 17: The use of Branch Software to do FB Product Linking 35

Figure 18: The Google Sheet to do Unbounce Linking 35

Figure 19: The use of Google Chrome Driver with Google Maps 36
API and the coding behind it.

Figure 20:The decision tree created using SPSS Software 37

Figure 21: The campaign strategy and the calculation 38

Figure 22: The Power Bi report for Grab-Kaodim Project 39

8

5. ​OVERVIEW OF ORGANIZATION

FreakOut Malaysia Sdn. Bhd ( The Appster Project):​

The Appster project of FreakOut Malaysia SDN.BHD is the mobile-first
digital media delivering, the latest mobile gaming entertainment to millions of users
and app developers and engaging with them through interactive communities in
South East Asia base. Its vision is to make all mobile gamer's fun, happiness, fever,
and interaction.​ ​This project aims to believe that programmatic native and videos are
the future of digital advertising. Its vision is to deliver exceptional benefits to both
clients and publishers, amplified the optimal engagement rates, delighted users by
offering websites and insert ads that look and feel identical. The organization
structure of The Appster Project is that FreakOut Sdn Bhd. is a company affiliated to
the Appster project. FreakOut is a marketing technology company, a part of
FreakOut Holdings Inc. ,which was established in October 2010 and headquartered
in Tokyo, Japan with a focus on the advertising and marketing technology industry.
FreakOut’s goal is to deliver unique solutions that enable brand advertisers and
publishers to succeed in their digital marketing initiatives. It’s mission is to create a
healthy and inclusive marketing ecosystem where each member of the adtech value
chain can thrive and co-exist. The industrial training program provided by The
Appster is to participate in all phases of the system development life cycle which
includes analysis, system design, coding, application testing, project deployment,
implementation, documentation, and support and to work closely with the rest of the
development team and to study and understand the user requirements and business
process to successfully develop the required system.

9

6. THE APPSTER WORK EXPERIENCE

6.1. OVERVIEW
In general, the working experience comprises into two parts: first is data

analysis which includes getting data, data mining, data editing, SQL query, python
coding and reporting it on confluence and second is system analysis that cover test
running the projects, reporting on Jira and scheduling.

It will be best described if it is started from elaborating the details of training
before the training started to be given. Next, an overview of software used during the
internship period, the nature of assignment and detail description on the data analysis
procedures and lastly describe each particular cases that have been handled for both
data and system analysis.

Figure 1: The above figure shows the nature of the work.

10

6.2. TRAINING
The training is divided into several types where on-boarding is come in the

first place. On the first day of the internship, brief and comprehensive on-borading
was given. This covers induction of the organization procedures, rules and
regulations. Besides, introduction to the manager, my supervisor and other team
members was given on this particular day.

It was short and informative orientation given by the human resource person
which also included the basic training on administration work such as the procedure
to send a letter, usage of photocopy machine, booking meeting rooms (for meeting
with the other team member sand my supervisor), email usage and procedure needs
to be taken during the office work.

Next, my supervisor is the one who assisted me on the first day to be
introduced to slack and opening mail, being added to the groups. He introduced me
to the softwares that were going to be used and the details of the projects which was
called ‘The Appster’ . This was given on the first week of my internship. Also i was
introduced to the ‘PUBG Documentation’ and how to use it as it was included in so
many tasks given to me. So the trainings basically covers the understanding of the
Appster project and the data that needs to be extracted and the process of getting
insights for the improvement of the existing app.

I got training API and testing the API links using different softwares and
exporting data from it and analyzing it. It was mainly done my supervisor but also
one of my team members were there to explain me in detail.I was also getting trained
about AWS (Amazon Web Service) & its multiple services like Athena, E3,
CloudFront etc by my supervisor and the assist.supervisor.I was also given resources
to know more about it and use it through my tasks. Besides, I was given training on
developing website and deploying it using new language for example Next.Js.

11

6.3. DATA ANALYSIS
The data analysis tasks basically started from my second week of my

internship. I was new to the software and the work environment so i was given
documentation to learn from. I was working on AWS Athena as i was given a task to
do a data query using PUBG API. I started with logging in on AWS and PUBG
Documentation. I did some simple data query on AWS Athena using SQL language
and showed it to the supervisor to give him an idea about my SQL knowledge. After
that, he introduced me to the ‘POSTMAN’ to test API. I had to use specific
commands to test if the link of the API is working well and if i can get data or no.

After logging in and i checked ‘PUBG Documentation’ to get the link from
the specific sector and with the id of the specific match that i need to analyze. Then I
put the link on Postman, added request type and checked if the status code is 200
which is basically a Postman response that indicates the response is received well.
After the response was positive i tried to export the file and the exported file was a
JSON file.I was not familiar to JSON files data so i was kinda having a hard time so
i did some research about it. After that I had to change the file to CSV or Excel file
to understand it more.

Figure 2: The above figure is the PUBG Documentation to get the PUBG
API links to get the data of the required fields.

12

After making it into CSV file i did research on AWS Athena and it took me
quite a long time to understand its working process as it was related to Amazon S3
Storage. So i had to research on both the services to analyze the data exported from
Postman. I tried to change the data format using Python but since my python
knowledge was almost a basic one it took a long time to change the format. After
successfully managed to do that i uploaded the file into Amazon S3 (Amazon Simple
Storage Service) storage as shown in​ ​Figure 2​ which is used basically to store any
amount of data. I had to make different buckets for different databases on Amazon
S3 so that it doesn't get clashed when i do a query for a specific database. When the
CSV file was added to S3 storage it automatically created the columns as it was in
the original database so it was much easier to do the SQL query.

Figure 3: The above figure shows the buckets Created on Amazon S3 and
uploading CSV files to each bucket.

I gave each bucket names according to the data so that later I won't be
confused with several tables. Each match had its different stats about different
telemetry events for each player. So it was really big data to work with and it was
getting updated every second or everyday as millions of players are playing PUBG
everyday. for example log player attack event, log player heals etc. So each time I
wanted to get data about any match I had to put the match id, get the links from
PUBG documentation, export the data and store it in Amazon S3.

13

Then I used Amazon Athena to query S3 data with standard SQL expressions
to get my required data as Athena directly connects to Amazon S3 storage and can
query data from the database. I had to specify the location where my data is stored
which is basically the bucket name on S3 storage and run it as it is shown on
F​ igure 3​.

Figure 4:The above figure shows query done on AWS Athena using SQL commands
from the stored database in Amazon S3 .

Using athena it was much easier to query the data as i the tables were
automatically uploaded on athena from S3. I just needed to specify the table name
and the column name i want to get data from. I could also join several tables and do
query from them. Most of the time the data used to be in JSON files as the exported
data from postman was in JSON files so i had to format it into CSV files first and
then upload it so that it easy to do the query on Athena and analyzing the data to get
the result.

14

On week 4, I started doing a new task to automate the data exporting from
API using python. But as I wasn't good at python coding that much my one of my
team members really helped me through the task. I used Google Colab which is also
known as Colaboratory similar to Jupyter notebook where we can code in Python
and write descriptions as markdown along. After getting familiar with these
softwares i started the coding part. I found it really hard in the beginning to
understand the language and use the correct commands but as it went I successfully
managed to fulfil my task. First I added commands to get the data using the API link
and then by adding the match id I got the match related stats successfully. My task
was to get the ‘mostly used weapon in the game by the players in four different
maps’. But when it came to getting ‘leaderboard data’ I encountered a problem. Each
leaderboard includes the top 1000 players for the specified game mode separated into
2 pages of 500 players each, numbered 0-1and it gets updated every 2 hours. But for
lifetime match stats i could only get data of 10 players with one request while each
page had 500 players. So I had to use ‘sleep’ command to get the full data. The data
that I was getting was in JSON file and it's hard to understand to analyze. So I coded
it to convert into csv files and exported the data much easily by changing only few
information like the matchId or the player name than doing it manually.

Figure 5: The above figure shows the codes of the automated process to get
data using API link from PUBG on Google Colab using Python language.

15

After I got the data I reported it on Confluence, which is collaboration wiki
tool used to help teams to collaborate and capture project requirements, assign tasks
to specific users, and manage several calendars at once. Then I did the retrospective
data analysis with my other data team and my supervisor and the data was correctly
exported as it had the expected result.

On week 6, I did research on how to improve the data getting process from
PUBG API without manually including the link everytime. I did research on what
technology we can use to automate the process like AWS Glue, Amazon Redshift,
AWS Data Pipeline, AWS Lambda, API Gateway, AWS Kinesis and so on. And
reported to my supervisor what technology can be the better suggestion.

I continued my research on the automation process. I also researched about
REST API. But it wasn’t the appropriate tool to be used. So I continued doing
research on the automation.

​Figure 6: The above figure shows the research result that I found which can
possibly be deployed to do the automation of getting data from API.

16

6.4. SYSTEM ANALYSIS
I created the project named ‘The Magnifier’ in JIRA Software and updated

the backlogs which is basically a list of tasks that needs to be done or those are
completed. I also added issues to a sprint so that the team and me can work on them.
Everytime new tasks come in I updated the backlogs as well as everytime tasks get
completed. Also updating the issues or sprint was one of my tasks. Besides, I also
had to do research on fixing website SEO link for The Appster.net link as it
encountered some issues but it was just a side task for me so didn't take much of my
attention.

Unfortunately, due to the fact that the company closed, I don’t have
any image or proof to show the tasks was done by me in JIRA neither on
Confluence. It was done on my 5th week.

At the same time, updating the backlogs and sprint was a regular task which
went till week 8. I have also done the system invalidation on a regular basis. But I
don’t have the image of the tasks done as I have lost the access to AWS after the
company got shut down. I have used Elastic Beanstalk which automatically handles
the details of capacity provisioning, load balancing, scaling, and application health
monitoring after uploading an application to invalidate the system manually as they
did not have it automated. I also used Amazon CloudFront which is a fast content
delivery network (CDN) service that securely delivers data, videos, applications, and
APIs to customers globally with low latency, high transfer speeds, all within a
developer-friendly environment. I was new to all these softwares and AWS services
so I had to do research on each service and go through the documentation to get a
clear idea and apply them in a proper way. My supervisor also helped me in
invalidating the system as i made mistakes on my first invalidation and it was quite a
mess. However, it got solved with the help of my supervisor.

17

Figure 7: The above figure shows the system invalidation process and scheduling
using Amazon CloudFront.

So as it is shown in ​ F​ igure 6,​ e​ veryday twice I had to the invalidation. It’s
basically once an article about any game gets posted on facebook or other social
media I have to invalidate the previous ad so that the recently posted article will be
updated. Also with elastic beanstalk i had to check the health of the application.
Basically, this is under the system analysis tasks i have done through my internship
period at FreakOut Sdn Bhd before it shut down.

18

6.5. WEB DEVELOPMENT
Besides data analysis and system analysis, I was given an extra task to make

a website or portfolio on local host using NEXT.JS. As i was new to Next.js I was
provided with the documentation and other resources to have a complete idea about
it to apply on my task. Basically, Next.js is a React framework capable of creating
applications that run both on the client and the server, otherwise known as Universal
JavaScript applications. I have used Visual studio Code software to do my coding
and deployed my website on local host. After deploying it on local host i used ZEIT
Now, which is a cloud company for hosting frontends and serverless functions. I
have used my HTML and CSS knowledge that I learnt in university for styling and
making the website. I tried to keep my website simple cause I was new to the
framework and the software. I checked some online website sample to have a better
idea on making a portfolio. Also, ZEIT Now made it really easy to deploy the
website as it connects to the Git integrations with GitHub to deploy an existing
project. This is the link of my portfolio that I made ‘h​ ttps://jasiatunazzina.now.sh/’​ .

Figure 8: The above figure shows the portfolio or website that i made using
Next.Js and deployed it using ZEIT Now.

19

7. SKILLS GAINED

7.1. PROBLEM-SOLVING SKILLS
As this was my first experience of the working world I had no idea on what

to expect. I have learnt a lot throughout my internship period. I have gained a lot of
skills including problem-solving skills, analytical skills, soft skills, data-presentation
skills, web-development skills and so on. Each phase of the internship has taught me
something new, I have gained new knowledge with each task completed. The tasks
were comprised of a lot of software that I have never even heard before so I really
faced a problem at the beginning but with the help of my supervisor and other data
team and with the help of google i have overcome most of it.

The main problem that I faced was getting data from PUBG API as I was not
familiar with API terms. But I was provided a sort of training by my supervisor and
given some documentation regarding this. After going through all of it I successfully
managed to solve the problem and got the data. I have also faced problems in
working with Amazon Web Services because it was one of the main tools used to do
most of the tasks including data analysis and system analysis. However, with time
and practice I became really good at it.

Another problem that I faced was when I started making my portfolio using
my Next.Js and deploying it using ZEIT Now. As i have never encountered these
terms before it was so difficult for me to understand the concept and create a website
within the given time.While i was doing the coding o faced problem using CSS in
Next.Js. But my supervisor helped me wherever I faced any problem and finally I
managed to create the website within the deadline. Also, I faced a problem in times
of system invalidation and accidentally I delayed the invalidation which caused some
issue to the article posting. However, I invalidated the system afterwards and it got
solved. Basically, I faced a lots of but came with the solution as well which helped
me to improve my problem-solving skills.

20

7.2. ANALYTICAL SKILLS
One of the main tasks was data analysis, decision making, problem solving

which helped me improve my analytical skills. I had to work with a large amount of
data as PUBG game is one of the most popular games in malaysia and the platform is
so huge which consists of a lot of data. The ability to take this large volumes of data
from its API and analyze its trends and understand the problem contributed a lot to
my analytical skills.

As the amount of data was huge, I had to decide which data to work with to
produce the expected result which helped me in decision-making. For example, in
PUBG API there are a lot of keywords but to get my specific task’s result I had to
communicate with my team members and choose the data keywords or field to work
on so that we get our result. It improved both my communication & decision-making
skills which are a part of analytical skills.

Also I have done a lot of research while I was doing the automation task for
getting data to come up with which technology should be used. So I gathered a lot of
information and made a data collect, store, scheduling process considering it the best
way to automate the system. Also, I have done data analysis using Python which was
quite a new language for me to use in the real world workplace. By updating the data
I managed to fulfil the needs and the task requirements and do the reporting.

I also learned data query which is considered as the first step of data analysis.
I have used SQL query, R language to produce the data on different platforms such
as Athena. Besides, I have also learnt about HTML, CSS, Next.js and so on.

21

7.3. SOFT SKILLS
During the three months internship at FreakOut Sdn bhd, not only my

technical or analytical skills got improved but also my soft skills have improved a
lot. The working environment was really friendly and nice. It helped me in working
in a team even though it was small. My communication skills have improved as we
needed to discuss together to do data analysis to decide what data to work on or to
analyse the data to get better results. Even though in the beginning I was quite a shy
person and didn't want to ask a lot of questions and tried to solve my doubts on my
own lately i changed. The friendly environment and my supervisors along with my
other team members made it so comfortable for me to ask them if I am having any
issue. I was also able to explain about any feature clearly to solve the problem. And
by discussing it always takes less time than solving everything on your own. So
working with this team really helped me to improve my communication skills &
teamwork.

Finishing a task within a time limit helped me in time-management skills. I
was able to finish all my tasks within the given time. Sometime it added a bit of
stress. However, it improved me to improve my crisis-management skills & helped
me to work well under pressure and not to panic.

My creativity skills also improved as i was doing the website designing and
styling, framing etc.Besides, it contributed a lot in my interpersonal skills
improvement such as: patience, tolerance, positive thinking, sensitivity and so on.
Basically, the internship period enhanced my employability, helped me to be
confident and provided me a real work experience.

22

8. OVERVIEW OF ORGANIZATION
Kaodim Sdn Bhd.

8.1. BACKGROUND OF THE COMPANY

Kaodim Sdn Bhd is established as a web and mobile platform that provides
an easy and dependable way for people to hire the right service providers. Kaodim
provides the ease of access to professional services like home renovation, logistics,
event catering, cleaning, health & fitness and other needs and promotes transparency,
accountability and improves the quality of services, making the market better for
everyone. The company was founded on November 3,2014 by two ex-lawyers
Fui-Yu Choong and Jeffri Cheong. Kaodim headquarter is located in Uptown 7
which has over 100 employees. In 2018, Kaodim became one of the top 10 startup
companies in Malaysia by Startup Ranking.

1.2 LOCATION
Level 9 Tower D, Uptown 5, 5, Jalan SS 21/39, Damansara Utama, 47400 Petaling
Jaya, Selangor.

1.3 VISION
Kaodim’s vision is to become the 'go-to' platform for just about any kind of

service that people need in the rest of South-East Asia.

1.4 MISSION
Kaodim aims to help SMEs grow their businesses persists. Its mission is to

empower thousands of service providers across South East Asia so that they provide

23

the best possible service - and at the same time, grow their businesses and
livelihoods like never before.
1.7 BRANCHES

Kaodim is currently operating in 4 countries which are Malaysia, Singapore,
Phlippine and Indonesia where the headquarter is located in Malaysia.
1.8 PARTNERSHIP
On 29/11/2019 Kaodim has partnered with Grab where users of Grab app can
request for Kaodim services through Grab app. This service is currently available in
Klang Valley and Penang only.

24

1.9 ORGANIZATION STRUCTURE
Kaodim consists of 13 departments that are Management, Malaysia operation,
Vendor engagement, Kaodim Singapore, Kaodim Phillipine (Gawin), Kaodim Beres
(Indonesia), Marketing, Talent, Finance, Product, Data, Technical and
Communications. Kaodim group is backed up by leading venture capitalists
including 500 Startups, BEENEXT, East Ventures, KK Fund and Venturra Capital.

Figure 9: Organizational Structure of Kaodim

25

9. KAODIM WORK EXPERIENCE

9.1 SOFTWARE COVERAGE
It can’t be denied that in this modern day, technology has been playing an

integral part in every part of human life. We are all agree that the vast growth of
technology has changed the trend of human workforce. Every industry including
human service companies in general has adapted with the changes by adopting
software to ease their work.

9.1.1. Analytical Tools: D​ uring my internship period I have used a lot of
analytical tools to do my tasks. To do data analysis and SQL Query I have used
Holistocs Software ​ , ​SPSS Software ,​ ​Facebook Ads Manager ,​ Jupyter Notebook
using ​Miniconda​ , ​Google Chrome Driver​ , ​Branch Software​ , ​Google Trends​ Tool.

9.1.2. Data Visualization Tools:​ ​To present the data or for data visualization
I have used several tools and software which includes P​ ower Bi ,​ G​ oogle Slides​ ,
Google Sheets​ , M​ icrosoft PowerPoint​ , M​ icrosoft Excel ​, M​ icrosoft Word ​,​ Holistics
Reporting .​

26

9.2. TASKS RECORDS
During my internship training as data analyst at kaodim Sdn Bhd, I was able

to work for many different projects. However, when I first started, I faced a lot of
issues. Earlier I was in a different company that closed down so it was again a new
start with new environment and work life. Since it was a big company I was quite
nervous and wasn't sure how to keep up with the tasks. However, , I was charmed
enough to have experienced hands-on and learned various parts of what is involved
in a project, how does a project run as per initially planned, generated and ended. I
also had the opportunity to be exposed to how much information and hard work is
fed in order to succeed in a particular project which made me realize the duty and
contribution of computer scientists in the world.

On my first day I was introduced to the whole Data team as it consists of
almost 10 people including my supervisor and was taken around the office to get
familiar with it. I received my On-Boarding email that contained information about
the company as well as the technologies used by the data team. I was also called in a
meeting with the team members and my supervisor to explain me about the
technologies and the database and the process or tasks i will be working on. So most
of my first day I spent on installing and preparing software that i will be using
through my internship. I was also given a Power Bi docs to look on and try to
understand the business process and a simple task to draw down my understanding
about the business so far.

My second day basically started with another discussion with the data team
members and my supervisor to explain them about my understanding of the first day.
After that my supervisor explained to me the business flow in more detail and the
problems that they are facing. The data team members also explained to me the
internal data and their main database. In the following days of the first week I had
meeting with the marketing team to understand the correlation between data and
business as in the data-driven business processes & how the marketing process
depends on data.

27

9.2.1. Vendor Sourcing using Google Trends
My task mainly started from the 2nd week of my internship. On my second

week, I was given a task to understand the customer relation management system
and the vendor management in a business company and do some research. My task
was to come up with a model to distribute the customers properly. Meanwhile, I was
explained about how the finance side work and the term GMV(Gross Merchandise
Value), GR (Gross Revenue), B2C rate ((Book to Completion Rate) at Kaodim and
how its calculated for each service. I also had a meeting with the marketing team and
helping them with vendor finding and other services that people browse the most
using Google Trends. So Google Trends was used basically to analyze the popularity
of top service search queries in Google Search across various regions and languages.
It is also used to see graphs to compare the search volume of different queries over
time. Though I was new to this website, I didn’t find it difficult as one of my team
members explained it to me really well.

Figure 10: The above figure shows ‘Google Trends’ to analyze the popularity
of different service search queries in Google Search across Malaysia for past 12
months and also to compare the search volume between different services for
example: House Cleaning Service & Aircond Service.

28

9.2.2. CRM Segmentation using RFM Score
After the Google Trends, I continued doing my research on CRM (Customer

Relationship Management) and my supervisor gave me some hints like analyzing the
RFM Score (Recency, Frequency, Monetary Value) based on the customer lifetime
journey or Curry’s Pyramid to come up with a proper CRM model and proof points
based on data. After research I decided to take RFM Score as my main focus to
create a CRM Tool or RFM Model.

However, to do the model I had to do a lot of queries using SQL on Holistics,
which is a Full-stack Data Workspace that is used for analytical works and also to
handle ETL/ELT, data modeling to data visualization, data exploration and delivery.
In the beginning I was struggling with understanding the database as it had so many
tables but by the time it became easier and with the help of my team members
understanding the database wasn't much hard. I spent almost like a week and a half
to come up with the model as I had to consider a lot of data to calculate the customer
contribution towards the business and to do the segmentation.On my 3rd week, as it
was the deadline for this task, I submitted to my supervisor. The mistakes that were
found and after several updating I successfully came up with a RFM model for the
CRM Segmentation and discussed with my supervisor.

Figure 11: Above figure shows the RFM model for CRM Segmentation that I made
based on data from Holistics and RFM Score.

29

Also, I came up with the promo or discounts should be given to the
customers based on the segmentation and the type of incentives. Besides, I have
done some data query requests using SQL came from marketing team on my 3rd
week as well and reported them back. I worked on vendor sourcing as well using
Google Trends to search for more vendors for some specific service types. It’s
basically web-scraping and updating Excel Sheet or Google Sheet with the data in an
understandable way and report back to the marketing team.
9.2.3. Country Analysis (The Philippines)

My 4th week started with the introduction of country analysis. All my data
team members including me was given responsibility to analyze each country. I was
responsible to do the country analysis for the Philippines. I was given a Power BI
Document to check each stats, how’s the business going there with the customers
and the vendors and what can be done. But only Power BI was not enough so i had
to do SQL query on Holistics to get the required data from its database.

Figure 12: Above figure shows the graphical statistics that i got after doing
the SQL query on Holistics for the Philippines and doing the Reporting &
Scheduling to check regular stats.

30

Figure 13: Above figure shows the findings and the hypothesis I got after
analyzing The Philippines data and proving the data corrected by proper reasoning.

After getting the findings I had to share it with my supervisor and explaining
the data team if there is any other suggestion to work on or to make it better. Firstly
it was done for a few specific services. Later on, I continued doing my research to
get more insights about other services. I was also given a data challenge by my
supervisor which was with another intern. It was a SQL query to do on Holistics to
get data and to have a better idea about Holistics which I have managed to finish
successfully and sent it to the operation team. Meanwhile, I have also done similar
analysis to the Philippines for Malaysia and Singapore. Because of the language
issue I didn’t really work on Indonesia. However, the country analysis was one of the
repeating tasks that I have done more than once.

31

9.2.4. Category Expansion
On my 5th week I attended a meeting with data, tech and engineering team to

do category expansion in Philippines, Malaysia, Indonesia and Singapore. I did
research on the cities that can be expanded in 4 different countries and the service
demand of the people. I also did research on the service types that can be launched
based on the vendor's existing in that particular city. It was a lot of searching in
google and getting the vendors were really stressful. Also i had to remove the
duplicate vendors. After finishing with the service types and the vendor details i had
to put them on Google Sheet so that the operation team will understand and can
contact the vendors easily. I also prepare interactive chart report on Power Bi &
Holistics by exporting data from Google Sheet and Holistics about the vendors and
services for example: how many vendors do they exactly need, how is the
performance of the existing vendor and if they need to be suspended and so on.

Figure 14: The above figure shows the report on Power BI which is the
incoming JR every month for different service types and for different business models
which was created by importing the SQL query done on Holistics.

32

Figure 15: The above figure shows the google sheet that I have made for
different cities and for different services which includes the city name, the service
types, the vendor details and it was done for four different countries.

9.2.5. Update data on Facebook Using Facebook Ads Manager

I was given a task to update customer list on Facebook by using Facebook
Ads Manager each month twice so that the campaigns that will be carried out won’t
be affected by the old data or old customer list. Ads Manager is basically a Facebook
tool to create and manage Facebook ads. It can be viewed, changed and checking
results for all Facebook campaign, ad sets and ads. Using the Facebook Ads
Manager was quite difficult for me as I have never used it before. But one of the
team members explained the process of how it should be done which made it easier
for me. Firstly, I had to do the SQL Query on Holistics to get the customer list from
the database and then upload the customer list on facebook using Facebook Manager
for 4 different countries. Once it gets uploaded successfully it is ready to be used to
run the campaigns and ads on Facebook based on the customer list.

33

Figure 16: Above figure shows the Facebook Ads Manager page where I
have updated the audience or customer list twice each month to keep the data
updated before every Facebook Campaigns.

9.2.6. Link Updating using Branch
The updation of links was done with the help of operation team. I attended a

discussion with them to get a clear idea of how to create the links for campaigns to
be used on facebook and on the Kaodim page itself. To do this task I have used a
Branch Software which is a data attribution tool to know the source of every
conversion or JR coming in for kaodim. I have done it for Facebook and the
unbounce page of Kaodim.

9.2.6.1. Facebook Product Catalog Link Update
Created the Facebook display Ads link using branch and implemented it on
Facebook to track the campaigns performance. To do that I needed to know the
campaign type, service type and what platform like Facebook or Google to ensure
the correct link on the correct product feed. I have done that for all the four countries
with the help of the data team as it was divided among us equally.

34

9.2.6.2. Unbounce CTA link update
Created quick link from Branch for MY, ID, SG, PH and

implemented the link on the Unbounce page which is the landing page for any
website like kaodim page. So I created both deeplink and quicklink.Basically,
Direction for the user is the deeplink and quicklink is connected to the app and
branch link is tracking link for each conversion of JR to know the campaigns.

Figure 17: Above figure shows the interface of Branch Software which I used
to create the Facebook Display Ads Link and implemented on Facebook Product
Feed.

Figure 18: Above figure shows the Google Sheet of Unbounce CTA link
which includes the deep link and the quick link for both IOS and Android devices.

35

9.2.7. Connecting Google Map API and Deduplication using Python
Doing the vendor sourcing for different countries was a repeating task and

really time consuming. I had to do it every once in a while. So to make it easy I tried
Python language to automate the process. Basically I, with the help of a few other
data team members, did Python coding on Jupyter Notebook, as I downloaded
Anaconda Software, that connects to Google Maps API and based on the input it
filters the vendors and the service type and collect the vendor detail like vendor
names, phone no, emails and so on. To connect Google Maps API we have used
Selenium which is an open-source web-based automation tool and used for testing.
We also downloaded Google Chrome Driver which is an open source tool for
automated testing of web apps across many browsers. Basically, after giving the
location and the service type name and how many vendors are targeted, it collected
all the vendor details and exported as a CSV file to work with it. The deduplication
was also done by using Python coding so that no existing vendor will be on the list.

Figure 19: The figure above shows the Python coding I have done on Jupyter
Notebook and the image on the right is the Google Chrome Driver page that goes
through the given location using Google Maps API and lists out the vendor details.

36

9.2.8. Creating Decision Tree using SPSS
I attended a training where my supervisor introduced us with the new

software called SPSS. After explaining its work process I was given a task to analyze
the complex statistical data. Basically, I used the SQL Script which was already
created on Holistics and got the data. After I got the data i uploaded it on SPSS and
chose the data fields I want to work with or I want to compare. Based on that i
created the decision tree and discussed with my supervisor my understanding.

Figure 20: Above figure shows a Decision Tree created using the SPSS
software and the statistical analysis of the data to understand the ratio more.

The SPSS software can also be used for the deduplication which I have done
using Python. But the difference between using Python and SPSS was it has an
interface while Python doesn't. However, I wasn't really comfortable with SPSS as it
was really complicated and I faced a lot of difficulties working with it. But with the
help of my supervisor and my previous knowledge from university about SPSS in
Probability & Statistics Course, I managed to finish the given task within the
deadline and created the Decision Tree.

37

9.2.9. Campaign Strategy Planning
I was given a Power BI document to analyze the Zero Quote percentage for

all the cities in Malaysia for Aircond service type. Basically, I checked the Power Bi
to know the exact numbers for Aircond Services, then checked which has the most
zero quote and what are the reasons. Because based on the reason I had to choose the
strategy if we need to onboard more vendors or we can check the existing vendors. I
was given a template by my supervisor to calculate and come up with the accurate
numbers of the acceptance request rate, active vendor rates and so on. After doing
the calculation I proposed the strategy that needs to be taken to my supervisor and
the CEO was also present in that particular meeting as it was for the campaign in
February and an urgent case.

Figure 21: The above figure shows the calculation of the statistics and the
campaign strategy I came up with.

38

9.3. PROJECT INVOLVED
9.3.1. Grab-Kaodim Project
Kaodim has a partnership with Grab where anyone who uses Grab can

request for a service using the Kaodim app feature there. This feature is only
available in Malaysia. All the service requests that comes from Grab are stored in a
different database and also analyzed differently. So this project has so many sectors
for example: vendor performance in Grab-Kaodim project Vs only on Kaodim,
checking the incoming JR through Grab Vs Kaodim and so on.

My task was to analyze the progress of different services and the vendor
engagement in Grab-Kaodim project . So I created the SQL query to know the
vendor engagement in accepting and fulfilling Kaodim jobs and compare with the
vendor engagement in Grab jobs as the Grab vendors are basically taken from
Kaodim vendors. Also I analyzed the Power BI document to check the regular ups
and downs and came up with the solutions of the problems encountered.

Figure 22: The above figure shows the Power BI that explains the vendor
engagement in Grab and Kaodim and the related statistics.

39

10. ​SKILLS GAINED (KAODIM)

10.1. PROBLEM SOLVING SKILLS
Data analysts are basically responsible for collecting and organizing data, and

then using that information to help solve problems or answer questions. And me
working in a position of Data Analyst at Kaodim, I had to do a lot of problem
solving. It also includes Data Mining for organizing data which doesn’t happen with
just the push of a button. Whenever any problem arises in an algorithm, a database or
lines of code, I was to be able to find and troubleshoot the problem so the data can be
efficiently captured. A lot of times for some wrong data in one data table the other
tables were not working well, so I had to check the data table and fix it.

Being a data analyst, I had to decide which problems to target and how to
solve it which really helped me in discovering my problem-solving skills as well as
improving them.

Previously, in vendor sourcing we were using manual way to search vendors
and listing them but it was really time consuming and stressful as it had to be done
often. So we came up with python coded automated system to connect with the
Google Maps API to do the task automatically and it took maximum few minutes.
Besides, I have also solved some urgent needed solution problems like the campaign
strategy for which I was given less than a day and i successfully came up with a
strategy to solve that specific service type problem.Thus, I experienced a lot of
problems and managed to come up with solutions which greatly improved my
problem solving skills.

40

10.2. ANALYTICAL SKILLS

Besides being a data analyst I had to do business analysis as well as its
related to each other. If the data is analyzed well it contributes a lot to business
improvement, same goes for Kaodim. The data team always works with the
operation team, vendor-engagement team and marketing team. Basically I had to do
a lot of analysis based on the data I got using SQL Query. I also had to analyze the
Power Bi docs to come up with an efficient solution for the business improvement.

Analyzing each country business performance we had to take up a lot of
decisions including on-boarding more vendors for specific service types or category
expansion for a specific city or country or city expansion in a specific country.

For running a campaign and to get the outcome also included a lot of
analytical skills. Each campaign needed different focus and that comes from proper
analyzing of the data. Thus, this position made me do a lot of business analysis and
data analysis which improved my analytical skills.

I also used different integration tools to collect data from different platforms
by using MySQL & do reporting. Each of them includes data exploration which is
looking to find interesting trends or relationships in the data that could bring value to
the business. So data exploration, statistically proving the findings, data analysis and
preparation altogether helped me in improving my analytical skills.

41

10.3. DATA PRESENTATION SKILLS
Data visualizations make trends and patterns in data easier to understand.

Humans are visual creatures, and since in a business company the improvement
mostly depends on data so the other t people need to understand the findings as they
aren’t going to be able to get meaningful insight by looking at a giant spreadsheet of
numbers. The data analyst position at Kaodim made me able to create plots and
charts to help communicate my data and findings visually.

It helped me in creating clean, visually compelling charts that helped other
teams like operation team, marketing team, vendor engagement team to understand
the data. I also understood the importance of avoiding things that are either difficult
to interpret like pie charts or can be misleading like manipulating axis values.

The tools that I have used the most to visualize my data findings was Power
Bi. Because I found it really easy to visualize data in an easier way and to explain
my findings to other teams to work on.

I have also used Holistics Reporting and Scheduling to make my report and
update it regularly. Basically the reports I made by using SQL query on Holistics and
exporting the required data.

Job requirements can vary a lot from position to position, but almost
every data analyst job is going to involve producing reports on your findings and
building dashboards to showcase them. I am fortunate to work with Kaodim which
enhanced my data presentation or data visualization skills.

42

10.4. SOFT SKILLS

10.4.1. Communication skills
Communication skills are very important when it comes to working in

the real world as it consists of a lot of employees, customers, clients and so on. As
Kaodim was comparatively large and only the Data Team itself consists of around 10
people so I had to communicate with a lot of people. I had to work with operation
team, marketing team along with data team which helped me to communicate with
them to share my own ideas and also listen empathetically to others.

10.4.2. Critical thinking

As a data analyst, I had to analyze situations and make informed
decisions. Whether I was working with data, operation team, or marketing team I
needed to be able to understand problems, think critically, and devise solutions. It
also helped to improve my creativity, flexibility, and curiosity. Thus, my critical
thinking skills have developed greatly.

10.4.3. Teamwork

I have improved my teamwork skills a lot during the internship
period. I had to work with a team of 10 people. Most of the tasks were divided
between the team equally so it took a lot of communication between the team and all
of us had to work with the same pace to execute a proper report. So each success of
tasks came with teamwork which helped me to improve myself in teamwork a lot.

43

10.4.4. Time Management

Time management was one of my challenging at the beginning of my work
experience where time costs a lot when it comes to the corporate life. Therefore, the
schedules, checklists and timetables helped me to improve time management and I
cannot deny that I am satisfied with my time management as I managed to finish
most of my tasks within the deadline.

10.4.5. Strong Work Ethic

One of the uniqueness of the Kaodim is its code of ethics and practice where
all the team including the data team working in a friendly and peaceful environment
with the mutual respect and productive cooperation among the employees in the first
place and between the company and its clients. Moreover, the willingness of the
team to help me with useful guidelines from the data helped me as well to gain more
practical knowledge and confident.

10.4.6 Self-Confidence
The work with Kaodim gave me the opportunity to grow better as an

individual and I have applied all the fundamentals and education that I have learnt
from my program’s subjects and practice them practically in a real work life
experience which resulted in a great growth in my soft and interpersonal skills which
lead to outstanding self-confidence and I am grateful for all the difficulties and
challenges that I faced who built up my confidence block by block.

44

11. S​ UGGESTION

Since there is no such thing as perfect, there is always room for
improvement. This recommendation comes in pure attention to a better future for all,
none were to accuse or point fingers nor to give mean comments. Hope the parties
involved will take action and make proper adjustments, or the least have
consideration to think about for a more exciting future.

As an intern, I was treated in a very impartial and professional manner which
made me feel pleasant and comfortable. Also, I am truly honored and grateful to
have been assigned to my supervisors here at Kaodim.. The only thing that I would
like to recommend Kaodim is to widen the working space for the data team as more
interns as well as full timers are getting hired almost everyday. So the office space is
keep getting consumed. Also, I would like them to keep the access to more learning
websites for everyone like Udemy, Udacity and so on.

I would also like if the university offered more workshops about resume
writing as well as interview skills. They can provide us with specialities to interview
us and review our resumes and give us feedback. That will improve our interview
skills and at the same time will make us experience it before the real time comes.

Last but not least, I would highly recommend learning as much as you can
when undergoing your internship as it is very beneficial for an individual to identify
their capability, weaknesses, potentiality and many others.

45

12. ​CONCLUSION

In conclusion, I believe an internship is an excellent initiative for
undergraduates. Based on my personal opinion, I consider myself very fortunate to
work with FreakOut:The Appster and Kaodim accordingly during my internship
period. Even though at FreakOut the experience was for a very short while it taught
me a lot of things and contributed a lot to my self improvement. At Kaodim, since
the office is very strategic and I have a good team who were willing to teach and
guide us with patience, it was immensely satisfying working there as an intern. In
overall, I think practical training is very helpful to students as it is a great stepping
stone for us to gain experience and learn to adapt to working culture before pouring
ourselves completely into the society. I enjoyed working in data team because of the
exposure of different types of tasks. It does broaden my knowledge towards different
kinds of data analysis or data science knowledge as well as business analysis
knowledge.

Besides that, six months period of practical training allows me to have a clear
view on roles and responsibility as a data analyst as well as gaining improvement on
communications, interpersonal and other soft skills. Most importantly, I was exposed
to the professionals' life of data analyst which affect me directly in keeping
professional in every aspect, especially as a data or system analyst.

I truly suggest every trainee and juniors who are about to undergo the
training, enjoy every moment of it. It really helps out a lot in the future. The
experience I have gained through the industrial training is valuable. Hence, little
piece of advice and recommendation, data analysis and system analysis are valuable
learning process, saviour every second of it.

46

13. ​REFERENCES
1. Freakout:The Appster Malaysia Sdn Bhd Website:

https://freakout.net/
2. Amazon Web Services Documentation:

https://aws.amazon.com/
3. PUBG API Documentation:

https://documentation.pubg.com/en/getting-started.html
4. NEXT.Js Documentation:

https://nextjs.org/docs
5. ZEIT Now Documentation:

https://zeit.co/docs
6. Kaodim Sdn Bhd (Malaysia) website:

https://www.kaodim.com/
7. IBM SPSS documentation:

https://www.ibm.com/support/knowledgecenter/SSLVMB_26.0.0/statistics_k
c_ddita/spss/product_landing.html
8. Microsoft Power BI Guided Learning
https://docs.microsoft.com/en-us/power-bi/guided-learning/

47


Click to View FlipBook Version