How to make a Data Science Resume - Tips by a Big 4 Analytics Consultant

I used to work as a data science consultant in one of the Big 4 Consulting Firms, with a master's degree in quantitative economics and B.Tech from IIT Kharagpur. Having conducted numerous interviews for data science positions, I'd like to share insights on crafting an effective data science resume.
It's not just about listing experiences, skill sets, programming knowledge, and data analysis expertise. Knowing how your resume is perceived, how it is shortlisted, and how it guides the interview process is invaluable information that can significantly enhance your chances.

A good resume should always be honest. If you include false information in your resume, a skilled interviewer will likely notice. I suggest being truthful, but that doesn't mean your resume should be empty. You can still add some interesting details to make it more appealing. First and foremost, it is crucial to define your role within the realm of data science. The data science field has three distinct profiles:

1. The Modeler:

   - This individual is primarily focused on the statistical aspects of data science, specialising in the development and optimisation of models.

   - Responsibilities include selecting appropriate models, determining optimal data sets, deciding on the number of observations, and choosing relevant features.

2. Data Engineer:

   - Specialising in the handling of data, the data engineer is responsible for tasks such as, filtering columns and rows, and managing ETL processes (extract, transform, load).

   - Proficiency in SQL, Oracle, and database management is crucial, along with the ability to collaborate effectively with data scientists to enhance the overall capabilities of the data science team.

3. Implementation Person/ML Engineer:

   - This role involves the practical implementation of models developed through the collaborative efforts of statistical data scientists and data engineers.

   - The foundational work in statistical data science and data engineering provides the basis for implementation, which includes deploying the models.

Having understood these things, talk about your education, add interesting points like certifications, courses, and projects related to data science, data engineering, or ML engineering. Next, discuss your internships and work experiences, emphasising those that match your target skills. If you don't have a lot of skills or experience yet, try to learn about what skills you need and relate your experiences to what you want to do.

Imagine you've created models but never deployed them, and now you want to show you're a good implementation person. Explore the implementation tasks carried out in the project. Then, describe in a few lines how the model was deployed. Remember, you don't have to be an expert in everything, just be honest and aware of what happened.
Avoid creating a resume by simply copying a format; craft one that is uniquely yours. I won't delve into typical issues with CVs and resumes, such as grammar, spelling errors, and accurate numbers, as these are implicitly understood.

In case an interview doesn't go as planned, interviewers often offer one or two opportunities for candidates to demonstrate their capabilities. Remember, it's not just your resume that secures a job; it's you who will contribute to the company. Demonstrating teachability and dedication can significantly impact your performance. Create a well-rounded, truthful resume, and focus on showcasing your abilities in the competitive data science job market. Don't think about what you lack; instead, emphasise how you can market yourself effectively.

Passing CFA Level 2 - Nine Mantras to Pass CFA L2 [Based on August 2022 Attempt]

In this post, I will tell you Nine Mantras for passing the CFA Level 2 examination. Those nine crucial mantras are mentioned below.

Cover the Syllabus 9 times
The first and foremast mantra is to cover the syllabus Nine times. Why nine times? There is a special reason behind it. The CFA syllabus is humongous. I am not saying that you need to prepare for the entire syllabus in a comprehensive way for nine times. You have to understand the length and breadth of all the topics and concepts in the curriculum. So, that you are in a position to relate to whatever is being asked in the examination. You do not have to take the pressure of remembering everything. You need to cover the entire syllabus as many times as you can and preferably nine or more than that.

Do not do Research on Quantitative Methods
I know quantitative methods is interesting, it has a lot of concepts and it is very demanding as well. You will be tempted to do a lot of reading on your own and try to understand that why certain things happen the way they happen and what are the assumptions and why are such assumptions used over here and what will happen if these assumptions are violated. Do not go outside of the scope of the curriculum. Quantitative methods encompasses a lot of things and with my experience of studying quantitative economics, I know that the content of quantitative methods is of two to three semesters. So, you have to understand that this is the kind of beast you are dealing with.

Corporate Issuers as the easiest topic
Corporate Issuers is the easiest topic, don’t miss it. Read everything thoroughly in corporate issuers. How can you miss something which is very easy.

Financial Statement Analysis (FSA) and Alternative Investments are text heavy
Financial Statement Analysis and Alternative Investments are heavily terminology dependent. So, there is very less mathematics and a lot of text. So, keep on reading as much as you can, but do not beat yourself up if you are not able to learn everything. 

Fixed Income, Derivatives, Portfolio Management
All these topics will consume most of your time, because they are heavy in terminology, heavy in maths and heavy in conceptual understanding as well. Do not take these topics lightly. You have to focus a lot in these three topics. 

These are the three pillars which you have to make strong- Fixed Income, Derivatives and Portfolio Management. 

Economics and Equity give the best ROI of your time invested
They will ask you to understand the concepts, remember the theory, and remember the terminology, but you will not find it difficult. You will find it easy and that will increase your score in CFA Level 2 examination. So, economics and equity are going to be your friends in your journey. 

If you leave ethics, you will cry
Try to read at least 20 to 30 minutes on a weekly basis, and it is not difficult, it is just requiring your attention. Do not leave it. 

Revise, Revise and Revise!
It means that you need to constantly revise the curriculum. So, if let’s say you have six months of preparation time slot, so revise after every couple of months. 

Calculator is your friend
Always know how to do any activity on your financial calculator, be it standard deviation, present value, future value, annuity, because in the exam, you will have to do a lot calculations and if you are not hands on with your financial calculator, you will hate yourself for it, because that is the silliest mistake you can commit. 

Why it took me 3 Attempts & 9 Years to pass CFA Level 2 ?

In this post, you will come to know about my journey of clearing CFA Level 2. Why it took me three attempts and 9 years to qualify level 2 of the CFA Certification Exams? Let’s discuss that.

Lost interest in CFA
The simplest and the easiest explanation is that I had lost all interest in the worth of CFA certification in my profile. I was into teaching. I was teaching physics. My career was not in finance. That is a big reason.

Apart from that, my 2 failed attempts were also characterised by my inefficiency or deficiency in the way of studying for this exam.

Schweser:
First inefficiency was complete reliance on schweser. Schweser is a CFA Institute Prep Provider. I completely relied on schweser for all the preparation and that is a suboptimal strategy. Schweser must only be referred for Revision and that too in the initial stages of the preparation because revision comes in two-three stages. There are stages of revision. The initial revision has to be done through schweser.

Problem Solving:
The second inefficiency from my side was solving a lot of problems. I practiced a lot of problems without understanding the concepts, that is not going to help you. Internalising the concepts and solving a lot of problems, both are needed. You cannot skip either of them.

Read a lot:
Third inefficiency was trying to read everything and remember everything of the CFA curriculum. My attempt to remember everything which was written in the CFA curriculum was a big mistake from my side. That is humanly not possible. The better approach is to get the essence of whatever is written because CFA curriculum is designed for people coming from different walks of life and all of them want to get into finance with the help of CFA certification. You need to cover each and everything written and prepare your concepts according to the type of questions which CFA is going to ask you.

So, before you get into the CFA program, try to get motivated for the program. If you are not having any worth or value for the certification, you will ultimately fumble. Even if you do not commit any kind of mistake in preparation, your lack of motivation will always be a hindrance.

 

Recession Proof Career - Economics Vs Data Science

Do this and you will always be in demand, no matter whether it's recession or it's layoffs. Do you want to be a data scientist or an economist? Do you want to study economics or data science? Let’s understand this with the conceptual ideas about these two separate fields and how they are different and how they come together. The rigour of data science is in mathematics and statistics. Obviously, it has been adapted for data scientists and data analysts, and when we are talking about mathematics and statistics, you can not discount the use of coding. At every walk of your data science journey, you will have to do some kind of coding. It could be very elementary coding with just one or two lines for commands, it could also be forming extensive functions and then calling those functions and then on top of it, you can also have the use of classes.

A skill which data scientists need to really be expert is in handling data, so when I say handling data, it means the file can come in any format and you must know how to make it in such a way that you are able to look at the data and find out some kind of elementary insights without doing any kind of extensive or rigorous modelling. Just take the data, clean it or process it in such a way that you are able to work with it. When you are a data scientist, you always have to worry about the business problem, about the business user, how your data science solution is going to help the business. When you are thinking about these things, about optimizing your code solving the business problem, the boundaries blur between software engineer and data scientist. If you are a data scientist, you will have a lot of members in your team who will be software engineers.

Let’s come to economics. When you are studying economics or you want to become an economist, there is a lot of rigour and assumptions about what exactly is the state of the world and how do you want to model that state of the world. You have to go through a lot of abstraction and then through the help of mathematics, statistics and your economic principles, you find a kind of solution. In data science, it’s slightly different. The more you abstract, the more difficult it is to link it with the real world and the more difficulty you will have in explaining your solution to the business user. Similar to data science, even in economics, you will have a lot of mathematics and statistics. Everything in economics needs to be actually proved. In data science, a lot of things would be just done with the help of showing, showing accuracy in charts. In data science, there are a lot of problems which are domain specific and sector specific. So, you become an expert in those sectors, if the need arises so, but in economics, a lot of problems have an underlying theme of political science, behavioural sciences, psychology, history, anthropology, sociology, a lot of social sciences find representation in economics. Also, there is a room for a lot of theoretical work in economics and you will get satisfaction in it. 

In data science, you have to support your theory with empirical studies always. You cannot just do theory. In data science, you always have to show your theory through some empirical work. If you study to become a data scientist or let’s say you have an experience of being a data scientist, you will be sought after by consulting firms, by IT or by any kind of a product firm, who are trying to build a solution or a product which requires data insights, predictive model, and some kind of recommendation, you get the drift?

When you are reading economics, or you are experienced in economics research, a lot of macro research firms like hedge fund firms, VC firms or simply risk consulting firms could go after you. You will be sought after, and if I just go up the value chain, if you are very good in economics, you have some kind of an advanced degree like a masters or a PhD with some amount of experience, you could work with central banks or big banks in their economic research teams, in their economist teams. 

In data science and in economics, there is a lot of mix. You work with data, you work with maths, and statistics, but the moment you take the economics route, you go towards a research route, you go towards R&D teams of banks, hedge firms, financial firms and the moment you go towards the data science route, you go towards IT firms, consulting firms and product firms. And yes, this is just a rough idea, I have heard of economists working with product firms and data scientists working with hedge fund firms. It’s not like that if I study economics, a lot of jobs will be closed for me and if I study data science, a lot of jobs will be closed for me. It’s just that build your profile according to your interests. And ultimately five years down the line, you never know what type of work you would be doing, because super-specialisation is going to be the theme for the next 15-20 years. The moment you equip yourself with a lot of super specialised and highly specific knowledge about a field, you will be selling like hot cakes and will go deep into the problem which you are solving. 

In this way, you will always be in demand, no matter whether it's recession or it's layoffs.

Best of Luck!

Data Science Course at Madras School of Economics (MSE)

Madras School of Economics is a very prestigious institute established in Tamil Nadu. The two year PGDM program at Madras School of Economics has two streams.

Research and Business Analytics
The specialisation in Research and Business Analytics offers electives such as machine learning, artificial intelligence, big data, and domain specific analytics courses such as medical analytics, in order to cater to industry demand in fast-growing areas. 

Finance/ Financial Engineering
The specialisation in Finance includes in-depth theoretical and empirical coursework in asset pricing and corporate finance. Electives offered include stochastic calculus, derivatives pricing, computational techniques in finance and market microstructures. 

In the first year at Madras School of Economics, the course structure is pretty similar for both the streams.  Only in the second year, the stream specific courses start kicking in.

If you look at the course structure, of the research and business analytics stream at MSE, you will see that it competes with the PGDBA course at ISI-IIT-IIM, Masters in Management (M.Mgt) in business analytics at IISc, Bengaluru, and Master of Science in Data Science at CMI. 

Course wise, it looks good. And the delivery of the course will definitely be open to subjective judgement. If you look at the placement statistics of the batch 2019-21, the highest of ~16 lakhs and the average of ~10 lakhs, with a 100 percent placement record, is better than the M.A. Program at MSE.

Becoming a Data Scientist at Indian Statistical Institute (ISI)

All the flagship Master's programs at ISI are essentially designed to be research oriented. Due to hiring spree in data science, these programs are very popular amongst the recruiters. You can easily get hired for a data science job, if you pursue these from ISI.

They are:

- Master of Statistics (M.Stat)
- Master of Science in Quantitative Economics (MSQE)
- Master of Mathematics (M.Math)
- M.Tech in Computer Science

There are other programs at ISI, which are more data science oriented. Let’s now discuss those programs as well.

- PGDBA (Post Graduate Diploma in Business Analytics)
PGDBA is the ideal competitor for IISc’s master’s program of management in business analytics. And in some way, it is much better than the IISc’s mgmt. program, because of the advantage that PGDBA has of three institutions, ISI, IIT and IIM. You will never get the advantage of ISI, IIT, and IIM in any other program as of now.

- MSQMS (Master of Science in Quality Management Science)
It is the two-year full time residential Post Graduate program offered jointly by ISI Bengaluru and ISI Hyderabad. The last semester is devoted solely for industrial internship. The course positions itself as a course which is meant to create data scientists. The placement is also very similar to other post graduate courses at ISI.

- M.Tech QROR
It is a very popular course of ISI Kolkata and specifically among engineers. This course is essentially about Statistical Quality Control, Reliability and Operations Research. If you want to know more about these terms, before data science was an established career, people in finance or people in operations research, were already doing data science. Hence, you can be sure that this course is also equally good, if you want to get into data science jobs. The internships or the final placements are pretty similar to what is offered to other ISI graduates.

- M.Tech in Blockchain and Crypto
ISI has recently started a two-year master’s program which is formally known as M.Tech in Cryptology and Security. If you go through the placement data, you will understand that as the flavour of ISI is in data science, most of the roles which are offered to M. Tech CrS students, are data science roles and that is going to change with the industry awareness for sure.

- PGDSMA
Post Graduate Diploma in Statistical Methods and Analytics is a one-year diploma course run by ISI North East Centre in Tezpur and ISI Chennai. This course is primarily designed to cater to jobs which has application of statistics and data analytics as their requirements. Placements are quite good, but not like the other master’s program of ISI, because PGDSMA is a 1 year post graduate diploma. 

Why I Quit my High Paying Big 4 Data Science Consulting Job

Everything was going great and streamlined.
I was the brain of the data science team at my workplace and that usually happens in a good company. When you have a talent, you get the recognition for it. And when I was working in PwC, I was a data science consultant there. And I was the mastermind behind all of the data science projects in which I was involved. My colleagues like Managers, Associate Directors, Directors, Partners, used to come to me and ask me various concepts and various application of data science and statistics. It was a very great phase of my career. 

Down the line, you have to think as to where this is going towards.
How this is going to be in medium to long term. When I thought of the medium to long term prospects, my career goals were slightly in different direction. I wanted a more senior kind of data science position, a more statistical data science position, a position which has more applications of economics and finance.

So, my projects involved a mix of data science, marketing, programming, business analysis, and much more. One specific project which gave me a lot of learning as far as 'R programming' is concerned, was a really game changing project for my career.
I learned a lot of programming just from that project. That project involved so many R codes, those codes were not simple, but like a puzzle. Once you unravel the key, everything starts fitting into place and then you understand the big picture and then you get the essence of the whole exercise and why it is done. So this was the scenario in my big 4 or PwC data science consulting job. 

The offer which I got after, is in a market research firm. The kind of work which I do in this firm now, is in the intersection of economics, statistics and data science. 

Everyone has their own personal journey. So, you might want to get into programming side of data science or you might want to get into the model development or statistics side of data science or you might get into a domain specific data science. None of these categories fit really well into my personality traits or my aptitude.
In my opinion, it is the mix of all these three and that is what really drives my career moves.
In future also, my career moves or project choices will be governed by that idea, that would refine my understanding of my expertise.