Professional Documents
Culture Documents
7 Step Ebook Guide
7 Step Ebook Guide
DataVader's 7 step
pack
“Data Science is the sexiest job of the 21st century.”
Either way, you are in the right place. Welcome to the 7-step
starter pack. The pack will help you get up close and personal with
Data Science.
STEP 1 STEP 2
What is Data Science? Different roles in Data
Science
Understand how Data Science
is the confluence of Get hold of different roles in
Mathematics, Programming Data Science and choose the
and business understanding right one for you
STEP 3 STEP 4
Gain the skills required Create your portfolio
Now when you know what role One thing that gets your foot
you want to work in, get a across the job door is your
road map of how to acquire the portfolio. With my definitive
required skills guide, learn how to make and
maintain your pirtfolio
STEP 5 STEP 6
Starting the job hunt Acing the interview
Once you are set to show your So now when you have a couple
Data Science expertise, it is of replies on your job
time to start with the job application, it' s time to
search. I have it covered for prepare yourself and crack the
you! interview.
STEP 7
Advance your career
Have you ever seen your mother predicting the monthly budget
with great accuracy?
While these practices give you enough intuitions to play with data,
you become a data scientist when you crunch, process, and solve
business problems with data. Basically, it is your job. Then, you are
called a data scientist
How do experts define
Data Science?
The definition I have given above is very basic and just for your
understanding. Let us understand how experts describe Data
Science:
According to IBM,
Did you get a similar question in mind? Why do I see so many ads
about a Data Science career?
So now you know why “Data Science is the sexiest job of the 21st
century!"
Is Data Science something
new?
It is natural to think that Data Science is a new discipline
altogether. Everybody keeps repeating “Data Science is an up and
coming field!”
As a data scientist, you need to code your way to get data from the
database and then clean, manipulate, visualize, analyze and then
share insights from the data
Now comes the million-dollar question
R or Python?
While you might initially be okay with such tools but as soon as
you would want more control over your data, you will have to
learn how to code. So better start it now!
Linear algebra
Matrices
Calculus
Let's say your manager comes to you and asks, why are our
employees leaving? The employee turnover rate at the company is
very high at 23%!!
It also means asking the right questions. For example, let's say you
are working in Netflix's marketing team. Let's say your company
asks you to bring in more customers.
You need to ask some questions whose answers can help you
increase your subscriber base.
Summary
Data Science is the art of gaining insights from data.
Know why you are seeing Data Science everywhere
Data Science is not a new field and it has recently come into
vogue due to an increase in data generation, computation
capacity, and development in packages and libraries in
R/Python.
Data Science is the confluence of three major components;
Mathematics/Statistics, Programming, and Business
Understanding.
Chapter 2
What are the
different roles in
Data Science
Is Data Science something
new?
In chapter 1, we understood that Data Science lies at the
confluence of three components:
Mathematics/Statistics
Programming
Business Understanding
They all are quite different from each other, require more
emphasis on one of the core skills, and also differ in terms of
value provided to the company. I will be elaborating on Data
Analyst, Business Analyst, Machine learning Engineer, Data
Engineer, and MLOps Engineer.
Let's say your company has set a yearly sales target of 1,00,000
units. The sales and marketing team wants input from you on the
following things:
If you are someone who is starting his/her career, you can orient
yourself as a business analyst or data analyst. They are in demand,
have a good number of openings. Also, the skills associated with
these roles are easier to acquire.
This job is good for an entry point, but if you want to code more or
do machine learning, then this role might not be the right fit for
you. Also, they pay less as compared to the role of a data analyst
or machine learning engineer.
But if you come from a non-engineering background, you can
think of getting into this role. Then gradually, you can become
better at programming, learn more machine learning and slowly
move to other, advanced roles.
Alongside Excel and other tools like Tableau, you will also need to
know SQL and database knowledge. In almost every data science
role, database knowledge and especially learning SQL is kind of
mandatory.
As you can see, the job requires you to create, support, and
monitor metrics to support business. It is also about delivering
metrics reports. If you are unsure what is a metric, a metric is a
measurable value that shows the progress of a company's
business goals.
Machine Learning Engineer
The primary job of a machine learning engineer is to develop
machine learning models which run continuously.
Apart from the models which you create, you are expected to
write clean, maintainable, and comprehensive codes. You will be
closer to the role of a software developer
Going back to the Amazon example, let's say you give real-time
updates on whether a customer on Amazon's website will buy the
product.
I feel like the role of Data Engineer is at a gold spot right now. The
demand is high and the supply of people with data engineering
skills is less. If you enjoy coding and managing data, this role could
be a good point as it is high paying also.
Time to analyze the job description of a posting for a data
engineer role at Google.
MLOps Engineer
Professionals working in MLOps help in the deployment of
Machine Learning or Deep Learning models and help to maintain
them. It means acquiring skills like creating web endpoints using
flask or fast API and also some cloud solutions like AWS or GCP.
The tools and techniques required for each role may differ and are
specific, I am going to talk in general on how to acquire the skills.
This chapter is more about the mediums and methods through
which you can gain the skills
No matter which one you go for, these are the must-haves that
you should be looking for:
Quality Content
Real-world examples integrated into the content
In-depth analysis encouraging free thinking, than just following
the steps
High-quality projects close to the industry practices
Guidance in building portfolios and cracking interviews
Option 1: Get an advanced
degree
Doing a master's degree or a Ph.D. is one of the options on the
cards. Doing a Ph.D. is a controversial opinion to discuss, so I will
focus on doing a master's degree.
The best part about joining a master's program is the network that
you build. The alumni network that you will access will help you
grow in your data science journey. You will study with peers who
are going to teach you a lot about teamwork.
The teachers will be like a guiding light to your journey and go that
extra mile to help you if you are in their good books. During my
applications, my teachers went the extra mile to write a LOR and in
guiding me through the process.
If you are not doing it from a good school, doing a job in the
meantime would be a better choice to go for. Some schools might
be having an outdated curriculum having less connection with the
industry that will do you more harm than good
So while choosing a masters
degree, be very aware of:
Cost
Time investment
The school and its
infrastructure
Faculty
Curriculum
Location (closer to the job
market, the better)
Option 2: Join a certification
program
There are several boot camps or certification programs being
offered by different upskilling platforms. Some with a job
assurance program and some come with a job guarantee program
asking you to pay the fee once you get a job.
In terms of investment, they will cost you a lot less. They are
cheaper than master's programs and they are finished within 6-12
months. The job guarantee programs enable you to take the
course, take a job and then pay later, which is a good deal to go
for.
You can also do these with your job, which is not possible with a
full-time master's program.
I get such questions a lot when I explore the options. The person
on the call would be throwing random courses and asking me to
pick one.
I will let you be the judge, I have written all the good qualities that
a course must have. You can add your personal preference to
them and then decide
Cons of certification courses
You will not have a peer to learn from and study along, which can
be demotivating sometimes. On top of that, the courses are self-
paced which makes it tough to schedule and complete.
You will not have an alumni network to rely on. The quality of the
content of these courses can be questionable, as they provide tons
of material and expect you to master them.
The assessment also does not motivate you enough to push your
limits.
So, as obvious, this a go-to option for someone wanting to get into
Data Science.
All the latest technologies and tools are taught on these platforms
at minimum cost to you.
Cons of Online Courses
Online Courses are a very good starting point, no doubt. But there
are some things which need to be fixed.
In a nutshell, they are good to start your learning journey but not
enough to advance your career
How is DataVader trying to
fix the issue?
I had mentioned above that online courses need to be fixed. For
that, I have come up with DataVader, which solves a lot of pain
points.
Projects have all the content that you will need. Even if you don’t
know the “D” of Data Science, you are most welcome, everything
will contain the basics.
With that, I would like to conclude this chapter. I hope you are in a
better place to decide which route you would like to take for
yourself. If you need any help, I am here
Chapter 4
How Do I Build My
Portfolio?
If you have reached this page, you are halfway done,
congratulations!
This is like a paradox out here. You need a job to get experience
and you need the experience to get a job.
By building a portfolio!
What is a portfolio?
In most simple terms, a portfolio is a collection of your works that
showcase your expertise in a particular domain.
Github profile
Kaggle competitions/notebooks
Blogs on medium, substack or WordPress
Personal website
If you are in a job and don't want to continue doing data science in
your free time, that's okay. That is your day job already :)
What is a project
A project is basically a sequence of tasks done to accomplish a
certain outcome. Data Science projects are similar.
They use different tools and techniques to get results and solve a
problem statement. With data science projects, you can
demonstrate:
Problem-solving skills
Proficiency with certain data science tools
Mathematical and statistical understanding
Ability to communicate
After you have done a project, you can essentially upload the code
on Github with an elaborate readme file. Create a blog post
around your solution, the challenges faced, and how you
overcame them.
One step further, you can even create a YouTube video to explain
it all. As we know, the best way to learn something is to teach one,
so teach, reach out to people.
Data Collection
You can collect data from various sources. You can run your own
survey, download some publicly available datasets, or you can
scrape data from the web. Also, there are APIs.
Data Cleaning
Feature Engineering
This is the step where you generate insights from the data and use
the data to derive actions that can be taken to improve metrics.
This I have discussed already in earlier chapters.
Modeling
Deployment
Now go and document your work as a blog and show the world!!
Explore as much as you can. Find a domain that interests you and
start doing projects. Don't only focus on state-of-the-art models
and the highest accuracy, also focus on documenting your work
and exploring the data properly.
Doing projects on the titanic dataset or the iris flower dataset will
not add much value as it does not help you distinguish from
others. Pick an interesting dataset.
How can DataVader help
you build your portfolio?
As we discussed, the portfolio is a very important aspect of being a
data scientist.
Data collection
Data cleaning
Data analysis
Creating dashboards
Modeling Deployment
These come with my 1:1 mentorship. I work closely with all the
participants and have a small group.
With that, I would like to conclude this chapter. I hope you have
gotten an idea of how to build your portfolio and land your dream
job. If you need any help, I am here
Chapter 5
How Do I Start The
Job Hunt?
So you have learned the basic skills, decided the role on which you
would like to focus. You have done a couple of projects and posted
a few blogs. The next step is to start your job hunt.
Before you start with the chapter, I just want to let you know that a
job search will not be easy initially. Gradually, you will build up
your network and land where you want to be.
Write your "About" in the first person, and explain your headline in
detail here.
Be polite, be kind.
You can mold your interview in the requirements that they have
asked which will help you gain an edge over other applicants.
This exercise will also help you to advance your skills by pointing
out the lacunae in your interview preparation. You can use these
guidelines to further upskill yourself.
Now comes the tricky part, reading the job descriptions, you might
get overwhelmed at times. Companies over expect from
applicants and mention a number of skills which is almost
impossible for a fresher or moderately experienced professional
to have.
In that case, you need to focus on your strength. Some HRs do not
know the requirements in detail and write whatever they find
online. Be patient and focused!
Using portals to your
advantage
There are a couple of online portals to search for jobs like
Monster, Naukri, Indeed, Timessearch, cutshort, etc.
But, I feel the lowest input and high return platform is Naukri.com!
If you are consistent on the platform and keep updating it, you will
get a lot of recruiter attention. Personally, I updated my resume in
January and I still get recruiter in-mails. Combined with LinkedIn,
Naukri has been my go-to platform for job search.
Now, I will not tag them as high ROI resource, but you source can
apply. Some of the companies might ask you to attach a cover
letter with your resume.
While this might seem like a draining and futile exercise, it gets
your foot across the door if written properly. To write a cover
letter, you can follow the I, YOU, WE approach.
While applying for companies, make sure that you are the right fit
for the job and the job is right for you.
Getting a job is not the end goal, you need to be aware of the
company's work culture and environment before joining. You also
need to look if you will get opportunities in the company to grow.
Let me know if you need more information and if I can help you
with something
Chapter 6
How To Ace the
interview
Till now, if you have followed this guide meticulously, you would
have received some interview calls. This chapter will help you ace
the interview, whenever you get the chance.
The same goes for the Data Science interview. As I said in the last
chapter, spend 80% time in upskilling and 20% time in applying for
jobs. 80% time should be allocated in becoming so good that even
if 20% of the time, fetches you an opportunity, you grab that one!
This means that you should be totally prepared for the interviews
in and out
This might seem daunting, you can ease this with practice! I am
going to include a couple of my personal experiences with
interviews.
Know what recruiters want
In the last section, I gave you some pointers over which a recruiter
might assess you in the span of time you spend with the recruiter.
Relevant skillset
There are two words to focus on here:
1. Relevant
2. Skillset
1. Take-home assignments
2. Live assessment
The second word is relevant here. It means that the skillset which
you possess should align with what the recruiter wants.
This will not mean that you have to go out and acquire each and
every type of skillset available. It means that you have to stay
focussed on openings whose demand for relevant skillset is what
you possess.
Business understanding
Managers have high expectations from data teams and when data
teams come with solutions, they are very complex which creates a
wide gap between them.
Team working
Case study: This section is to check your data skills. You will be
provided with a dataset and will be asked to manipulate data,
visualize, create a machine learning model, depending on the
job you have applied for.
You have to be really good with your data manipulation skills using
different Python libraries like pandas, matplotlib, sci-kit learn, etc.
Problem
Statement
Your
approach
Impleme-
ntation
Tools used
and why
The
impact
There can be more interview rounds, depending on the company
and its management.
The prepared answers will give you confidence and reduce the
randomness of the process.
This guide can be elongated further but I will take a pause here.
Chapter 7
Advance Your
Career
Once you get the job, now comes the final step!
The first thing you might want to do after you join the job is to
meet the manager and set the expectations right.
You may feel that you should already know the job expectations
from the job posting and interview process. Although this is
sometimes true, a lot can change between the interview process
and the start of the job.
The interviewers may not be in the same time frame as you, or the
organization may have changed before you joined. By talking to
your manager as early as possible, you’ll get the most up-to-date
information and have time to spend discussing it.
You can’t know when you start a new job what the expectations
are in terms of job responsibilities.
If your company has been doing data science for a while, a great
place to start is by reading reports that employees have written.
Reports will tell you not only what types of data your company
keeps (and give you key insights), but also the tone and style of
how you should communicate your results.
Then you’ll need to learn where the data lives and get access to it.
Getting this access includes knowing what table contains the data
you want and maybe also what data system has it. Perhaps the
most frequently accessed data lives in a SQL database, but the
event data from two years ago lives in HDFS (Hadoop Distributed
File System), which you need to use another language to access.
Making the job change
You can think of applying to different companies once you have
spent a bit of time in your first company.
If you have had some complications in your first job, like a toxic
work environment, unsupportive managers, or anything else, you
can think of shifting jobs before that as well.
While choosing jobs, don't make salary as the only criterion, think
of fit. Whether you are fit for the role or not.
With this, we come to an end to our eBook. I hope this eBook was
helpful. if you'd like to know further or join DataVader, you can
visit datavader.io or mail me at datavaderio@gmail.com