The epicenter of ‘How to build a good portfolio, for your Data Science interview’ is GitHub which should have at least 2-3 strong projects.
A good Github profile is not a quantity game but a quality game and if you want to have a look at any specific repo, you can just click on the link here and refer to one of the GitHub repos built by our student, which we believe is a good representative sample of what a good repo for a Data Science project should be.
A good repo typically has a very strong cover note, it has a very strong analysis plan, the presentation that was put together with respect to when you portrayed the results, it has a summary of what is the final recommendation, it has the whole iteration plan of what all that you did, and what worked for you.
Presenting a complete understanding, the whole journey perspective of what you did on a particular project is extremely important.
The actual algorithms to find the parameters that you suggest that need to go in the final scores with respect to the metric that was chosen all of these things are necessary, but I think a good project portfolio repo is a more complete document, it is not just a piece of 500 lines of code, it is something that really helps you get started and showcase your thought process which is very important.
We’ve seen really, really strong portfolio projects where people have gone on to create some sort of a Loom video, or some sort of a complete analysis node, or some sort of demonstration, especially if you’re doing a lot of stuff on the Computer Vision or Natural Language Processing, so that’s how good repo looks like.
With respect to what should be the spread of topics in your portfolio, we definitely recommend one strong project, which you have done right from accumulating the data to deploying it, and by deploying it, we mean that you at least have built some sort of a user interface where you’re showcasing the functionality, or at least, opened up an end-point which people can access and put together some sort of query and get back some sort of a payload or some sort of JSON and that’s very important that you’ve done a whole journey end to end and have deployed a project is very important out here.
With respect to the sort of topics, we would typically say that just make sure that you hedge your bets, you have a supervised project, an unsupervised project supervised learning project, an unsupervised learning project, and at least one project in Natural Language Processing and that gives you a fair amount of variety, which is going to cover at least 70% to 80% of all the interview types, or different topics that the recruiter may expect from you from a portfolio perspective in the interview.
Nonetheless, going through the journey, you would have identified, what is your strength, if your strength is a particular algorithm if your strength is a particular area of study, then we just recommend that you go all out and bring that aspect out very strongly in your GitHub repo.
GitHub these days allows for a lot of cover pages, it allows for building many websites using Jekyll, so max out your experience on GitHub so that you’re able to portray strong yourself through a strong portfolio.
So there you go on GitHub, you can post your repos, you can post your blogs, you can post your complete website, and everything that is essentially required to present yourself as a strong professional.
Let’s spend the next one-half minute to understand what does a Data Scientist typically do on any given day.
What I’m just going to do is break down this into some key activities that practicing Data Scientists have done or do on a full-time basis.
There are multiple stakeholders that Data Scientists typically works with.
Let’s take an example of a person who’s working in a services company where you have a client to cater to, and you have a customer possibly geographically in a different geography, and their data and so on and so forth.
So the number one thing that you will be doing as a Data Scientist is that you will be accessing the data from the customer from their live environments or remote environments, and either bringing it to your environment if they permit to already be completely operating out of their environment. A large part is all about making sure that the data is sanitized in a particular structure and, you know, it’s there ready for the analysis? What we typically can call pre-processing, which can involve a lot of cleaning imputation, just making sure that it is sanitized and it makes sense.
The number two thing is you’ll be interacting with the client just to first understand that what are the data or what is the data? What are the table headers for it? What is the business context that is assigned to it?
Next will be understanding typical business processes that govern the creation of that particular data as if it’s an inventory management system, then how does a particular record get reflected on what does it really mean? All those Nitty gritties that you typically get into.
The next thing that you would typically be doing is trying to understand what is the business problem that your client is trying to solve? Are they trying to cut costs to grow their revenues, or they’re trying to curb attrition of customers or trying to understand the underlying behavior, or trying to disclose some sort of key drivers. These are just a representative list, but essentially, you try and understand what the problem is.
Once you have the data and you have assigned the business connects to it, you have understood the problems, what you do thereafter is essential, break that problem down from business problem to a Mathematical Problem, which can either be automated, or you need to apply some sort of Machine Learning Algorithm to its mapping and come to a conclusion that this is the problem and this is how we need to solve it which is part of the self-learning process. Post, which you put together is some sort of a solution plan and go through it in a very iterative process.
All of these things don’t happen in a single day. It’s more of a phase and a journey but, at some point in time, you’re going to come back with a certain analysis, some sort of exploratory data analysis, you’re going to present your findings back to the team and they would give you feedback that you know what you are on the right track, you’re thinking about a business in the right way, in order to go ahead and put together some sort of solution. Once you put together the solution, once you come back with your recommendations, you again, go back to the client, and showcase your findings to them.
This phase of the journey where you’re actually taking the results from Machine Learning, which are metrics and some sort of matrices and some sort of scores and you go back and deliver some sort of business segmentation and that’s when the client again, goes back listen to you and there is a slight bouncing off ideas and essentially come back with a recommendation which you ultimately give to the business. This is what happens in a typical services based organization, it can also be with respect to the client be replaced with an endpoint or service or some sort of web app, which is essentially delivering some sort of recommendation and you’re constantly improving it and putting your recommendations in action and are getting some sort of feedback, which can be a multi-week cycle, there are various workflows.
The important part is a Data Scientist is a person who understands the business getting into the tech, understanding the core craft, that is solving problems with data, and then presenting it back solving a problem, iterating through it, it’s a journey that goes on and this is how a typical day of a Data Scientist looks like.
It’s a very interesting day, you will not be working on the project solo, you’ll be working with a team. So you have a deep understanding of a particular algorithm or a particular domain or business side, and then other people on the team who can add different flavors to it. So it’s a very happening place, a very interesting journey.
It is a very exciting career track that’s in the offering for you. If you need to discover more about how to create a Data Science Career Track, click on the download brochure button right here, and I’ll be very happy to share with you all the other data points to help you get started.
Actually, there is a larger question over here where people typically from a lot of non-technical, non-engineering backgrounds or people who have had phobias of programming or maths, this is a question that comes from them because they believe that they are self-selecting themselves out of the race, you know, they cannot become a Data Scientist with typically a question that’s coming from that population.
So let me just dispel a few myths over here.
Anyone who’s done 10th, 10+2, and if you’ve paid some decent amount of attention to what has happened over there, I think it’s really helpful because I think there are some foundational skills that are being covered in that journey.
A lot of us during that stage and phase of our lives do not look at some of the topics like statistics and maths from a very applied perspective but what matters is that you walk through the grind, and you will be able to revise and, develop and solve those same problems, again, but a mathematical intuition, which is far more important.
So I would say 10+2 is the basic minimum requirement, post which if you have done Bcom, BCA or any other 10+2+3 kind of a degree, which is non-technical in nature doesn’t matter, I think you’ve done your 10+2, that’s really awesome.
Secondly, from a mindset / mental perspective, what we definitely recommend is that you have a deep interest in the business side of things where you want to solve problems for people. So the ability and interest to solve problems are very important. What do you want to do and what you’re going to learn as your core skill is that you’ll be solving problems for businesses, clients that you will be working with, with data, you’re not going to be solving problems with technology, or you’re not going to solving problems with algorithms. I think the key the crux of it is that we will be solving problems for your customers with data. Of course, technology, and algorithms play a very supportive role but problem-solving is a larger environment with data.
I think the third part I would say is a personal trait, which I would say is curiosity. Where I think the ability to question things ability to deconstruct a larger problem and break it down into smaller chunks, essentially, unending quest of being wanting to ask questions, is super, super important.
The number four thing that I would very strongly rate over here is the need the ability to learn continuously, I think you will need to keep on learning, there is no fixed set curriculum that you will exhaust and you will get a job and that’s it, you’re done.
I think you need to keep on learning because I think the amount of information that just keeps on happening in the space is huge and you just have to keep on learning picking up new stuff, and you know, making the move as the market moves.
So there you go. So these are some of the things are required to become a Data Scientists loads of enthusiasm, curiosity, wanting to solve problems, basic 10+2 and you can make a very strong Data Scientist out of yourself.
The larger question over here is — Is there is a magic set of tools that we absolutely need to know, before you kind of crack the Data Science role.
What I’m going to do over the next one minute is just going to help you understand two things.
One is, I think, one of the tools that we typically cover in the GreyAtom way of doing things, our curriculum, and more importantly, what are the various topics that we typically cover in the career track at GreyAtom.
So number one, I think Python is definitely the language of choice, we are very well invested in Python for multiple reasons, you can check out a separate video for that pattern, a being a general-purpose programming language, skill-able, supports object-oriented functional paradigm, so on so forth. So definitely, it’s something that is a language of the future and a simple google search a Stack Overflow report, with respect to one of the leading languages in 2020 and beyond, Python is definitely going to show up over there.
The next thing is getting at the library level, we’re definitely interested in NumPy, Pandas, a lot of scikit-learn, for NLP we are very much into NLTK, and a whole host of other libraries, which we keep on picking up for specific things. Like if you’re doing scraper that there is, you know, critical soup and a lot of other things that we do pick up from time to time. And some of these things also change, you know, batch to batch code to code. So if you find something that is relevant, interesting, we just pull it in, for that particular batch.
Come into the second part, which I promised is ‘What is the broad flow of curriculum at GreyAtom?’
So we start with the absolute basics, programming, What is Python? How do you start programming? The scientific part of Python, from there, we want to stats, we want to the basic treatment of maths that is required for Machine Learning, we get into the exploratory data analysis, model building, a lot of regression classifiers that are typically getting built by a wide array of algorithms, and there is a very long laundry list of it and finally, we want to Natural Language Processing we also have a precursor a small introduction to Computer Vision. Plus, I think there are topics that see you some treatment with respect to time series with respect to chatbots, so these are topics of some potential because if you get into one of these interviews, you need to have surfaced some basics in all of them.
I hope this answers some of your questions, you can definitely download the Data Science right here
Let me answer this not with a number that you know what there are 32,000 jobs in the market right now, at this point in time, the answer that we’ll offer is very, very simple.
What I would encourage you is to go to a portal, which is representative of at least the domestic market that I would say is Naukri.com or Monster.com, not being the flag-bearer in the space, why don’t you go to Naukri.com and have a ready list of designations that you may want to quickly search for or just put Data Analyst, Data Scientist, Data Science Engineer, Machine Learning, I think once you start reading some of these keywords, you will start getting a number and this number is typically going to cost around anywhere between 20,000 to 40,000 depending on what search term you put.
Now the thing overhere is that some of these roles are listed by the companies and also by some of these agencies and the middlemen in the whole job or procurement market. So the way to do this is that, you know, take the number and take at least 70% of that as the true value.
So there you go. I hope this helps you discover, you know, what is the number of data science jobs in the market, and that should be anywhere between 30,000 to 50,000 at any given point in time.
One last thing to note over here is if you looked at this number, not just yesterday, today, but over the last 12 months, you will see that this number is not just staying where it is but it’s actually constantly growing. irrespective of how many 2020 has been, especially with the food situation, the demand for the Data Science role has remained extremely robust.
So I hope this answers your question and please feel to reach out to us if you have any more questions.
I’ll just help you understand What is the GreyAtom way of doing things and this is definitely not a question that can be answered in 60 seconds but at GreyAtom we recognize this as 40% of our journey, what we typically call Career Services.
At GreyAtom we have a very deep understanding of how this process of hiring typically works at a very contemporary version of it.
Right from preparing the resume, a resume that will pass the human test, a person will able to see the resume for six seconds, and shortlist the resume, plus a resume that will also bypass and clear the bot that is installed in the application tracking software that your recruiter is using.
Next, having the right social presence with your GitHub profile with your LinkedIn profile, making sure that it is engineered not one day before the application starts but for at least three to four months duration so that your profile looks very very strong on the day of application.
The third is your portfolio, your data portfolio, make sure that you have a good mix of projects that are reflected in your portfolio and this is not a number game but this is about the quality game, the kind of analysis, the depth of analysis, the way you’ve presented your explanations, all the things that work for you all the things that didn’t work for you a very accurate representation of your thought process.
Number four is having a very strong presence on the blog, we recommend at least anywhere between four to six really strong blocks over a period of six months, which really help the other person understand that you know, you put in the effort, you have a deep understanding, so on so forth.
And from there on, I think these are just you know, something that you need to have read even before you start the application but once you start applying, once you start preparing for an interview, it can be specific to a company, it can be specific to a particular domain or field of study.
Secondly, I think it can be specific to a hiring challenge. So I think it’s a very fairly neutral catapulting.
A couple of resources is one is that you should check out the YouTube channel of GreyAtom, we will definitely find a lot of resources, specifically addressing this question that is How to prepare for a Data Science Interview and you should also stay subscribed to Instagram handle of GreyAtom because we do a lot of these sessions or at least a couple of times per week, where we speak about the Career Services aspect of Data Science and also the Front-end engineering program.
So let me just quickly refresh your memories that there are two career tracks that we have created.
One is in Data Science Engineering and the second is in the Front-end Web Development kind of a role.
Both of these career tracks come with the option of paying 75% of the fees once you get a job and the first 25% that you need to pay, you only need to get started with INR 10,030/-. So that’s the short answer to the question over here.
How do I get started? How much do I need to pay?
There you go it is INR 10,030/- , the whole structure designed in a way which keeps you focused which keeps you motivated and which keeps us jointly focused on the mandate that is to get you a job.
Remember, there is only one definition of success at GreyAtom and that is when you become successful, and that is when you get a job.
Now, the thing over here is the underlying question is that a lot of people are fearful that when they start their careers, especially when they switch their careers, from being a programmer, or any kind of techno-functional role, they’re really apprehensive about whether they will need to start off as a complete fresher and their experience will be disregarded.
So guys, the companies, the recruiters, the hiring manager, they value your experience, they know that you’ve gone through atleast three, four appraisal cycles worked on a particular domain, you’ve solved some business problems, all these things are extremely valuable for the prospective recruiter.
Secondly, you actually represent the majority, we find the overwhelming number of people who are in the four to eight years experience packet and even at this point in time seeking to transition into Data Science Career.
To directly to answer your question, how much will you get paid, any recruiter, any hiring manager, any employer or company, they’re going to look at your current compensation, they are going to understand what is the commensurate compensation for the role on offer and basis of whatever is the market condition if the market condition is very bullish, and everyone is getting like 20%/30%/40% hike, then you should expect something very similar.
All of these things come down to not what your previous role experience or anything is, but how strongly have you come across in the interview process? How is your portfolio project? How was your hiring challenge? Did you explain what you did very, very nicely? So all of these things matter or everything else?
I think it is purely skills, it is about demonstrated excellence that matters and nothing beyond that.
So how much will a 5+ years experienced person get paid in a Data Science Role?
The short answer is your current compensation plus whatever is the prevailing market uptake that everyone is getting in any given market.
Well, let me start off by saying, and dropping a few names over here.
Varun Nangalia, I think he is one of our most recent pass outs and he is actually a Chartered Accountant by training and now a Data Scientist.
So yes, it is absolutely possible we have a lot of people who are working in government, pseudo governmental agencies, accounts, sports, psychology, sales, marketing, and all the different fields, who are actively training to become a Data Scientist.
I can just pull a lot of these examples, who are essentially professors in engineering colleges who went on to become Data Scientists, people who were into pure sales, door to door sales, technical sales, pre-sales, who went on to become a Data Scientist so all of these are possible.
The reason is, there has to be a proven step by step method, which guides you on how to tread the path, how to make this transition happens, someone GreyAtom has a very deep understanding of how this transition really works, how to make it happen, consistently, repeatedly and that’s exactly what we do.
Having a very good understanding of stages, we understand that when you start, we start with absolute basics, and we know that you know, I think there’s a chance that you actually run away and upright and give up on your dreams in the first 20% of the journey. So that’s why it is a very high level of handling, post which we start building your concepts, so that’s where you have a lot of freedom self-paced, a mix of video interaction, live mentor training, then you need to get into teams and groups and start solving problems start solving into and projects. So you need to get into hackathons, at some point in time, you will need to craft a custom strategy for yourself your roadmap to getting a job. That’s where some peer interviews, mock sessions, and everything come in.
What I’m coming to is that irrespective of your background, you can make this thing happen and we’ve seen it happen enough times to say this with a lot of confidence that this is possible and you can make it happen. There’s no significant disadvantage or advantage that comes if you’re from an engineering background, anything of that sort.
If you’re an active programmer, at this point in time, I would say you definitely have you know, some heads up, we definitely have some, head starts but again, it’s not something that is later sustained or sustainable advantage to have at the start that you can capitalize on it but although there’s no need to get discouraged, you can make it and we have no evidence of people having made it.
Do check out our GreyAtom Success Stories page and you will find many more such examples of people having made transitions from all sorts of backgrounds.
There are some very well documented ways of finding internships, some of those conventional approaches may not always work.
So here are a few tips that I’m going to share.
Number one thing is, use LinkedIn to find internships to find hiring managers but don’t ever reach out to someone and say, I need an internship, I need a job and ask them as a cold message that you sent on LinkedIn, this doesn’t work.
What you can do is you can actually build out a blog post or you can put together a Jupyter Notebook and ask for feedback and reach out to people and say, you know, why don’t you review this? I need some feedback. Can you help me, there is a higher chance where people may end up volunteering to help you, this is just one of the tips on How do you really reach out to the people, so on so forth.
Beyond LinkedIn, there are places like GitHub and they have some specific features with respect to pull requests with respect to issues with respect to queries, with respect to how do you reach out to people on some of these other platforms like, you know, GeeksforGeeks and lots of different things.
So just make sure that whenever you’re reaching out to someone socially or otherwise, this is a very strong context, which is based on competency, which is based on trying to build something trying to ask a specific query and starting a conversation should be a priority rather than asking for an internship role or a job. So that’s essentially the thing.
So how do you find an internship or Data Science internship, of course, you know, go through your normal usual suspects exhausts that list, apply on portals on so forth, but use some of these things and use them enough. Apply the law of large numbers, reach out to multiple people use all these innovative techniques and you will have success for sure.
The short answer to this question is that you can have the access to the brochure by clicking HERE .
What I’m going to do next for the next 30 seconds is just help you understand how is this curriculum built, how it is iterated.
The GreyAtom curriculum is built with feedback from multiple stakeholders.
Number one, are the companies that are hiring and we are constantly talking to them and then refining this curriculum.
Number two, we are also talking to stakeholders from the academia, specifically the IIT Madras, and they give us a lot of inputs with respect to what should be in the curriculum, and what can be possibly taken out or improved.
Number three, we always keep on talking to our students, people who get placed our alumni past, present, and all the students.
So that’s how the curriculum is typical, you know, built and iterated.
Now the curriculum is built by the people who are actually Data Scientists, people who are working in the industry, and they have actually built up this curriculum is, of course, the instructional design team that’s in house, the way it gets actually built, embedded integrated with the product that is all taken care of.
Lastly, the curriculum and the content is not a static entity. It is constantly evolving, there is a state where it is rapidly getting built and it is rapidly getting built again and again. What we have today is possibly the fourth version of the Data Science content, not just the curriculum content we have built, we rebuilt and rebuilt it again. So that it is something that is up to date, and something that is required for getting a job.
Finally, with respect to designing a curriculum, it is actually a problem of plenty. What you really need to focus on is to take away what is not required, and just keeping on removing things with only a very small critical masses remaining, which really matters to all the recruiters and which helps you crack the job.
Yes, absolutely. They are important
Data Visualization and the people who essentially come back from a career where they have done a lot of BI tools, a lot of business objects, some sort of Crystal Reports, report genital MicroStrategy, or any of these drag and drop tools where you do a lot of slice and dice data, you build dashboards, CXO level dashboards is very, very important skill.
A lot of people essentially do find a lot of jobs and do something that also has a sustained demand, specifically Tableau and at GreyAtom we have a module that we also train people in tableau, especially the executive curriculum.
So if you are doing something like Python, so there are a lot of these libraries, which will help you visualize you know, whether even if you’re doing your EDA for something like as simple as Matplotlib, or Seabourn, and I think if you just most likely a few steps ahead, you’ll find a lot more mature libraries that will also help you build very strong, you know, dashboards, something like a pixel plotting library, like highcharts, or d3js, which are not exactly visualization libraries, but they’re more pixel plotting in nature. So they are important, it’s a very important part of what a Data Scientist is, and being a Data Scientist.
However, it’s not the core. It’s a very important part. It’s a very important subset, but it’s not the core. Being able to communicate is very important, but it’s not the core.
So if you’re just choosing or planning to get deeper into a tool like Tableau, and you’re trying to build a career on top of it and wanting to make a progression, or planning your future benchmarking hedging against that, probably it’s not the safest of the ideas, you might want, you know, slightly more broad-based set of skills, more programmatic capabilities, and then you want to be moving at it.
So yes, the short answer is Data Visualization important, absolutely important.
Do people get jobs? Yes they do get jobs, but there is something far more fundamental that enables you to visualize something that you’ve done prior to that, and that is the code and this is a very important part, but not the core.
GreyAtom is headquartered in Mumbai.
We are based out of Mumbai.
Apart from this, we also work in multiple countries, namely, we have a batch cohort that is working and running in Canada, we have a presence in Sweden.
We are also running batches and training for companies in the BFSI sector based out of the UK and we also do a lot of training for companies which are based out of Pune and Chennai, our office is in Mumbai.
But the bulk of the operations, the training and the students and the footprint that we have, is global and is totally completely pan.
At different points in time our team keeps on traveling to make our training happen across locations, get the feedback or it talked to a lot of learners everywhere.
Our community is global and has multiple chapters across different cities.
So yes, we are global in that sense, which absolutely international and we have a very, very massive global footprint.
The batches that we conduct at GreyAtom are online, they are remote, they are live mentor led, so you have one mentor.
But how many learners do we have in a batch, and that’s very, very critical.
Here are a few things that you need to know.
As a learner, your experience is very, very personal, all that you’re able to see is a mentor is a screen that has been shared by the mentor and the code alongs that are happening over there, and possibly, you’re able to see a chat window there, where the more the number of learners, the more the will be the variety with respect to the questions that pop up and those are the questions that you may not be able to anticipate those are the questions that are not coming to your mind but those are the questions that are coming to the mind of different people and they’re asking some really interesting questions.
Your learning is accelerated because of those questions, so it is actually important to have those people in class.
Number two, I’m sure you’re going to have questions at different points in time.
The way you ask questions in these online classes is actually to type out these questions and when you type out these questions, you’re actually crystallizing those thoughts in your mind and you’re actually formulating a well-formed question and it actually forces you to think and when you put out those questions, there are always going to be follow-up questions by the people, by peers and how the whole interaction happens.
So how many learners do we have in GreyAtom Batch?
The answer is typically around 80 to 120 but the reality is more the merrier.
There’s no student-teacher ratio, the student-teacher ratio is very much a relevant concept when it is a classroom and is a physical classroom, where there are some restrictions with respect to how many people can you accommodate in a class and there is a question of maintaining discipline within the session, the whole decorum piece of it.
In an online environment, things are far more structured, you’re able to mute, unmute, allow people to annotate, annotate it yourself, share screen share multiple screens have breakout rooms. The more the merrier is the reality
Once again, how many learners in the backs 80 to 120 is the answer but remember, it’s always good to have more number of people in the class.
Let me keep this thing really, really simple for you? The answer is a strong NO.
The only thing that matters in the interview process is whether the employer feels that you will be able to solve their problems and if you demonstrate ample evidence of that with respect to your resume, portfolio, your performance in the hiring challenge how we are able to find your projects when you’re focusing and talking to them in the whole thing, that’s the only thing that matters.
No certification, no degree is going to assure of anything with respect to cracking a job in the Data Science.
So take it in writing from us, we have seen close to 1200 plus people transition after their stints at GreyAtom and this is one thing that we have a very very good handle on the reality and the class.
The answer is NO.
There is no bulletproof degree, Ms. MTech, anything that’s available over here. Focus on skills, focus on demonstrated skills and that’s all that matters.
2020 has been a very challenging year .
Given the situation how do you increase your chances of getting a Data Science job?
I just want to highlight a few behaviors that our learners, the successful ones, repeatedly continuously exhibit these behaviors and this is something we have identified from Data.
We are observing and have quantitative evidence of people who log on to our platform and they come back and the are able to showcase that you know, this person is succeeding because of this reason.
Number one is consistency learners who keep on coming back to the platform every single day, for 60 minutes minimum, every single day, 60 minutes minimum, they have a significantly higher chance of being able to make the transition happen for them.
Number two learners who come prepared for the sessions by completing the pre-work that is mandated for that particular session, they again, stand a very good chance to make the most of that session and be able to apply it very, very strongly.
Number three, is when the learners complete their hackathons with all the sincerity, and they complete it out, and they do, you know, multiple iterations of it, that’s when they have a significantly higher chance.
Number four, I would say is when the learners do the assignments and projects with sincerity.
In the GreyAtom learning platform, we know that the learner has attempted a particular assignment once, twice, thrice, four times, five times, and then has taken the hint or has taken the solution, or the learner has got it right in the 10th attempt but did not take the hint of the solution. This tells us something about the resilience of the grit of the learner.
So all of these things increase your chance of getting a Data Science job, exhibiting the right kind of behaviours, showcasing it from day one, building it, making it your character.
That’s what makes a very, very bright use case for you to make this transition.
This is a very frequent scenario that happens that people go through five failed interviews and suddenly they end up converting not just one offer, but multiple offers.
They were chasing the jobs, and now jobs are chasing them because they have just got the knack of how to go about tracking interviews very frequently.
The question that comes very often is that, you know, I have these two or three job offers, what should I select?
Let me throw light over here.
Number one always select a very strong mentor, a very strong team over compensation. Just make sure that you’re associating with the right set of people. And you’re going to work on a very interesting project that is going to expand your horizons, with learning with knowledge, and there’s going to be a lot more growth over there.
Number two is the domain, there is no one domain that is, you know, a better world than any other domain. It’s not that the farmer does some interesting work that is happening. And there’s no work happening on retail or credit cards, or finance or BFSI or anything of that sort. But something that personally excites you, if you have that kind of clarity, awesome, go for it.
Number three when you’re upskilling yourself, you possibly would have solved multiple projects, done capstone projects, and you have some sort of an inkling with respect, you know what this is the area that I really want to go deeper into, this is what I enjoy. This is what I really want to specialize in Natural Language Processing or a specific problem within Natural Language Processing. That is your calling, go for it.
So if you have a couple of offers where you in one place, you’re able to do a lot of NLP based and the other place you want to do a lot of fraud analytics, credit card, that kind of thing, maybe NLP is something that you need to pick.
Finally, if you’re getting to do something that is truly cutting edge that is has a very strong rarity value, it is going to be very, very unique in a limited time span. Go for those kinds of profiles goes for those kinds of roles.
So there you have it, value, the mentor value, the domain, value, the company, value the kind of work that you will be able to do value the rarity of work that you will be able to do.
Don’t place unnecessary premium with respect to what is going to be your first compensation after you upskill yourself because the money will come. The right compensation will come but just focus on doing some good work.
Data structures and algorithms are a very fundamental topic and a very popular topic, especially for the web development roles
The important question over here is that do we need data structures and algorithm when you’re applying and trying to crack a Data Science interview.
Data structures and algorithm are a very essential part of your programming career, everyone should know about it, everyone should have the experience with it and this is the most critical part of your journey but this is not something you know, if you don’t know, then you will be stuck at someplace, this is going to improve by your experience and once you start learning Data Science, essentially you will grow with the data structures and algorithms.
The thing with data structures and algorithm is that there these things that are tossed around the big O notation, time, space complexity, you know, how to derive a particular notation, a lot of sorting algorithms, and algorithms in general.
Now, the thing is that, yes, some of these things and their understanding and their application, it grows with time when you’re actually solving some of these specific problems.
Don’t try to cram them in your mind to begin with, whenever you’re learning any programming language, whenever you start, you inevitably end up learning a lot about the data structures, what are the various constructs in any language? How are they supported?
So these are necessary, however, no one is going to stop you. These are not going to be your stumbling blocks, so to speak with respect to cracking an interview.
Having a fair entry level understanding of all of these things is absolutely necessary. But just make sure you allow the understanding of this to grow on you organically over a period of time, which has its foundation on application experience of having built things of having done things, rather than picking up a very thick textbook called as Thomas Cormen and going through the 1000 pages, or trying to get into some sort of a crash course anywhere else.
Don’t cram it, make sure that it grows organically, in you with you. It’s important, but it’s not a prerequisite and it won’t hinder or stop your progress but it will just enable you and help you grow into a larger professional over a period of time.
The answer to this question is a big YES
Just like cricket, in life placement, timing is very, very important to get the right kind of outcome from your shots.
Does GreyAtom help its learners fetch the jobs with the placement and all that happens and that is needed?
It’s a very central question to our philosophy and as I said, for GreyAtom there’s only one definition of success and that is ‘Your success is our success’, once you get a job, that’s a successful event for GreyAtom.
So yes, getting the learner placed is of paramount importance, and the way we typically do this is we dedicate anywhere between 40% to 50% of the duration in the main flagship program, to help you make that kind of transition.
This can typically be in the form of making sure that your resume is good, your GitHub profile is good, your LinkedIn profile is good.
You’re given enough practice with respect to mock interviews, peer interviews, with respect to how to solve problem statements and Data Science problems in the interview setting, how to solve hiring challenges, all of these things which contribute towards you getting placed is part of the GreyAtom’s Career Services and we can speak about this at length.
We speak about the human aspect of it how to keep yourself motivated when you’re going through a spate of resignations how to increase your odds of getting shortlisted and this is a science in which we are mastered.
So yes, GreyAtom helps the learners not just once for a period of one month, two months, but it’s a lifelong process that we are creating have embraced to help our learners out.
On a very serious note, we have also helped learners not just to get a job for the first time but also to help them seek a transition and in certain cases to also help them find another job in case there has been an unfortunate loss of pay, even for them.
The answer to this question is a big YES.
When we tell the recruiters that this resume is really strong, this person is really good. They trust us.
You know why?
Because that learner is GreyAtom Endorsed.
Let’s look at it from an industry perspective, let’s look at it from an employer’s perspective.
Their problem is not that they don’t get enough resumes, they’re getting enough resumes from the market to screen but the problem is that they get 100 resumes and they are able to shortlist two and the time it takes to reject that 98 resumes is exceptionally high, you just compound this problem to a different magnitude and you have a very different problem.
GreyAtom when it goes to a company with 10 resumes, the company will be able to shortlist 7 out of them and that’s what being GreyAtom endorsed means
Let me just break it down simply for you.
You have the GreyAtom learning platform, you are logged into the platform, we know when you’re logging in logging out whether you’re copying your solutions, whether you’re solving it on your own, whether you’re struggling through it everything we know.
Now basis of your quantitative performance basis of your attendance basis of your qualitative metrics, like the amount of time that you spend, how you approach the problems and your behavior.
We are able to churn out a score that tells us whether you will get a job, how soon will you get a job? What is your level of preparedness?
We know before you know that you’re going to get a job and that’s what GreyAtom Endorsement Framework is all about.
Once the GreyAtom learner is endorsed, we’re very happy to share those resumes with our industry partners and typically such candidates move really fast and they are lapped up by our partners.
So there you have it. That’s what the endorsement framework is all about.
Trophies, certificates, certifications, they play a very important role.
If you’re trying to get into a college on a sports quota, it’s important. They help you in your academic life to give you motivation, support at the right point in time, it’s a very strong incentive system.
Whether a certification is going to help you get a Data Science job, the answer is NO.
Let’s look at why.
What is it that the company is looking at the company is trying to onboard you to see to check whether you are the right person to hire whether you will be able to solve their problems. That is the number one thing that the company is trying to hire.
Globally, we have seen a trend where companies are moving away from grades, certificates, pedigree, your background, your job, and gap in education, and more on the hiring process. How did you solve the puzzle? How did you solve the challenges and whether you can do it on a sustainable basis going forward in the future?
This is also reflected in the fact that a lot of even these MNCs are now moving to a national process.
Let’s take the example of TCS or KPIT, where they’re conducting nationwide processes and not campus to campus processes. Of course, the campus to campus processes exist but there is also focused on nationwide processes.
Number two is that Data Science does not come with a set curriculum that you need to exhaust. It’s not like the AWS four levels of certification, or the Cisco certified. Basically, all the Cisco certifications that we have, there is no standardized set curriculum that you need to exhaust to fetch a job.
Every six days, you have a fresh library, a new take a new implementation that’s coming in the market, and you need to be on the top of the game.
So certifications per se, don’t play a very key role and they don’t dictate and they don’t help you define yourself.
If you’re using a certification, to stay motivated to stay focused or anything of that sort is good, but other than that they don’t really serve a purpose.
Finally, when you aim for certification from any institution body, academy course, upskilling agency, what you really need to understand is that when you’re paying your fees to them, you’re also paying part of your fees to the university with which they have an affiliation. So your money is actually going to go and be paid as branding fees for some of them.
So just make sure when you’re doing a program in Data Sciences, emerging technologies, you’re investing in the knowledge you’re investing in the skills you’re investing in experience acceleration, but not in paying the branding fees for certain agencies.
So the short answer to this question is NO, certifications don’t pattern and they don’t help you get a Data Science role.
help you get a Data Science role.
Gone are the days when companies flocked the campuses, and they picked 200 students from a single batch to put them into IT roles.
For Data Sciences, the hiring process a little different. The shift that we see, and this is globally and also very well reflected in the Indian market is some very, very strong trends are emerging.
Number one, competencies, skills demonstrated excellence over certification, grades, marks, pedigree, and colleges. If you’re able to showcase that you are able to solve the problems that the recruiter has, you have a super awesome chance to be able to make this thing.
Number two, there is a very level playing field, even the campus hiring process is now significantly flattened out due to national processes, so TCS is conducting a nationwide process, the KPIT is conducting a nationwide processing and multiple such things. Even they have realized that they need to not go to their usual suspects with respect to tier-one colleges, but also tap into the pool, globally and nationally.
Number three, the process, the hiring process is now definitely where they are doing hiring challenges, where they want you to attempt the problems that they have in their company in a small way, and come back to them in a thought, you know, with your thought process with your approach, and have a discussion on top of time.
Number four, is definitely your portfolio, your GitHub repo, and everything that you have built together with respect to your social presence on LinkedIn, on Twitter, it makes a lot a lot of difference.
Of course, a strong resume, a resume which will clear view test by a recruiter by an HR and which will also clear the bot, a typical modern-day application tracking software has, which typically just scans your resume gives out a score where you know, what is the matching percentage, so on so forth, and just shortlist automatically or not? All those things.
So having all of those things, you know, a one-page resume a cover letter or resume that covers the human and the machine and, you know, crosses them meets the requirement, having a strong portfolio, cracking the hiring challenges, being able to handle the HR process, the leadership round, the stress test, the negotiations, all of these things are absolutely necessary with respect to hiring and this is how our hiring process typically looks like.
There is no one set template or process or method but typically you should expect anywhere between a 3 to 6 step process with any of these companies.
We first need to understand what is the hiring process?
What do you think is the recruiter trying to look in you when they’re hiring you.
The recruiter essentially has a business problem that they want you to solve, they want to get confidence in the hiring process that you will be able to solve the problems that they have in a meaningful way.
The way to do that is to take a larger problem, take a smaller subset out of it and they essentially give out to you to solve over a period of two hours to two weeks and see what you come back with.
What they’re looking out for is not a complete, foolproof reduction grid solution but what they’re looking out for is your approach.
How do you apply your mind?
What kind of questions you ask?
What kind of solution you essentially come up with?
A hackathon, or a hiring challenge, which is done in a classroom environment may be remote, maybe where the students are distributed, mimics the same mechanics very, very closely.
This is something that builds the essential muscles in you to be able to solve the hiring challenges that are given to you in a real-life environment when you’re going for the interviews, even if you’ve never seen the problem, even if you don’t know how to solve it but how do you methodically approach a particular problem deconstruct it and solve it meaningfully?
That’s exactly what a Hackathon is and that’s exactly what is a Hiring Challenge.
At GreyAtom, we have this very, very well documented, it’s almost a proprietary method of making the learners go through it to assimilate their knowledge, to bring everything together from solo individual concepts to application of it.
These hiring challenges and hackathons are essentially where the learning is accelerated and your problem-solving skills come to floor.
That’s what is a hiring challenge and that’s exactly what is used in the hiring process these days.
2020 has been a year of many surprises, COVID-19 being just one of it.
Is the Data Science job market impacted? How does the future look like for Data Science jobs?
Well, here is the real dose of reality at GreyAtom, in the last seven months, and we are starting in December 2020. In the last seven months, we have outreached to more than 4000 companies. We have had a lot of students that are getting placed in GreyAtom for Data Science, we have in fact had the best months with respect to the placements.
So is the Data Science job market impacted?
The fact of the matter is that programming has become a foundational skill programming is something that is going to be taught to a six-year-old, your Data Science is going to become a foundational skill that is going to be absolutely necessary at the undergraduate or graduate level.
So the demand is robust. There are a lot of jobs that are getting created and will continue to be getting created in the Data Sciences space.
The future is really, really strong bullish and just make sure that next time you are ready to ride the wave rather than getting swept by.
So the short answer to the question is YES, the job market is rocking. It’s booming and it’s going to be like a brighter side early in 2020.
like a brighter side early in 2020.
There are multiple schools of thought on this one, do you need to learn programming to become a Data Scientist.
There is this school of thought, which is called as low code, or no code, which is essentially a lot of drag and drop tools, which will help you do very, very effective Data Science, I have nothing against them, to be very honest but it is our school of thought and it is my school of thought it is GreyAtom school of thought that we believe that a learner to develop a lifelong differentiating competing edge, they need to develop programmatic capabilities.
What they mean, what I mean is that you need to have at least one language that you’re mastering very, very well and you’re able to, you know, move on with that in a very, very, very meaningful way.
So do you need to learn programming to learn Data Science?
The answer may be NO.
You can possibly do without it, especially with a lot of modern-day tools that are available but should you, you know, we kind of put it in the must category that you absolutely need to pick up a language and be able to do.
The reason is, it is not only clients that you work with, but sometimes you work with systems, with endpoints with API’s with systems, which are beyond your control with services, so in the modern-day, you know, web application framework, it is absolutely imperative that you have strong programming capabilities.
So the answer from GreyAtom from me is 100% YES, you absolutely need to pick up programming to be able to do it.
This is a very popular myth especially among professionals with 4-10+ years of experience that when they do make that transition, they will actually start off as a fresher.
We’ve seen enough number of people make the transition, even with 8 years of experience, and they don’t need to start off as freshers.
The way this works is to put yourself in the shoes of the hiring manager of a Data Analytics or Data Science company.
They are going to value
Number one, the number of years of experience that we have done, it means that you have the organizational experience, it means you’ve won multiple rounds of appraisals, you know how companies work, you know how clients work, you know how business works. That is absolutely valuable.
Number two, if you have been working for the last 10 years, you’re definitely solving some business problems in some domain, you’re working in retail, you’re looking in finance, you’re working in CPG, you’re working in pharma, you’re working in healthcare, you are solving business problems, and you have an understanding of the business side of things.
And number three, at some point in times, you will definitely interact with the clients handled technology, so you have the third additional domain knowledge that advantage from the technology side or something to that effect.
All of these things are valuable.
The way your compensation is going to be structured is whatever is your current compensation, plus a hike which is commensurate with whatever is prevailing in the market. If everyone is getting 20% / 30% / 40% hike, then you should expect something in that range, depending on how well you come across in the interview process.
So the answer to this question is NO, you don’t have to begin as a fresher there is NO Restart.
It’s just a step up from there and it’s continuity and it’s a continuation.
As a child? I remember a lot of people, you know, my peers asking, what is the use of maths? Is it ever going to be useful at any point in time in my career is going to help me get a job?
You know what? Maths can actually help you get a job.
But if you’ve hated maths, it doesn’t matter. It’s okay
We’ve see a lot of people who for one reason or the other, ended up hating mathematics at some point in time and now have this question in their mind — Can I become a Data Scientist?
What is more important is not the math, the derivation, the sums, the formulas, but the mathematical intuition and what I mean by that is, it’s not what is the formula? Whether you remember that or you don’t remember that but do you know how to apply it? Do you know why to apply where to apply how to apply?
With a lot of modern-day tooling, languages, libraries, you may not need to do the heavy lifting, but you definitely need to understand the underpinnings of what’s under the hood, how it’s working.
Now, here’s the second most critical part over here a lot of people make math the pre requisite, it means that you need to know math before you get started with programming before you get started with Python before we get started with Data Science and Machine Learning and that’s just plain wrong.
You need to have your bearings in maths, absolutely, that’s necessary but that is something that you can develop over a period of time, you can start attacking the subject from the programmatic front, you can start learning Python, you can start solving some assignment, some problems, some projects, and once you get deeper into it once you’re actually solving the business side of things, and then you kind of discover the underlying mathematical complexities and nuances. maybe you will have an everlasting and a more firm and solid understanding of the mathematical concepts.
So two key things one, you need to be applied maths, we need to intuition and number two, it’s not a prerequisite it is something that we can build over a period of time and we have seen both of it is absolutely possible with all the students that you’re working with, whether they are from engineering, non-engineering, non-technical background, it is 100% possible.
At GreyAtom, we have only one definition of success and that is, your success is our success.
When you get a job, that’s when we define a successful event has happened.
So there is no pseudo criteria like completing a programme or getting some sort of certification.
We’ve linked our success to real life, real-world outcome and we’ve also linked our financial success to that.
What we mean by the pay later program at GreyAtom, the career track is that once the student gets a job, once the first salary comes into the account of the student, that’s when the student pays 75% of the fees to GreyAtom.
In the beginning a student has to pay an admission fee, a registration fee, which is different for different programmes and he can get started.
All you need to do is just pay a total of INR 10,030/- and get started and the 75% of your fees is paid once you have a successful outcome.
We conduct the 5-day workshop regularly and have done this for close to 8000+ learners.
Here’s the single thing that we’ve learned.
Most of the people, 60% of the people who we talked to have never coded, or they have coded 10 years back. So that’s just a plain and simple fact of life.
So if you have never coded, can you become a Data Scientist? And then that’s the question.
My response to that the answer to that is YES.
A language like Python, which is a general-purpose programming language, is not more than 900 minutes away from you 30 minutes, every single day for 30 days, 30 minutes, every single day for 30 days, that is 900 minutes, you will reach a place in your python competency, where you’re able to chase your larger dream of becoming a Data Scientist, onboarding concepts and learning a lot of things.
So I would say, please spend a good amount of MONTH, 30 minutes every single day for 30 days, you will be able to make this kind of a transition.
It’s not too difficult. It’s like English. It’s in fact very, very easy, very intuitive and you can do it you’ve seen enough number of people who have started from absolutely scratch and have done this.
One last thing we at GreyAtom acknowledge this and when you get started with the Data Science with Artificial Intelligence program at GreyAtom we start with absolute basics like what is programming we start from that particular point on?
So the answer to this question is, YES you can become a Data Scientist and there is a proven scientific, repeatable way of making this happen.
The question that students really want to ask is, how long will it take? Before I get a job in Data Science?
Let’s talk more about GreyAtom’s flagship program in Data Science, which is called Data Science with Artificial Intelligence and how long is the program?
Let me answer both questions.
Number one, the program at GreyAtom is for 8 months, on average, out of which a good 4 to 6 months is just spent on upskilling you in getting all the skills necessary for you to get that job-ready within you and we focus at least 2 to 4 months, between those eight months, just to get you ready on the career aspect on helping you get that job.
So by the time you are in the fifth month, you will actually start applying and you will start going through that grind of cracking the interview.
The short answer to your question is 8 months.
And the short answer to the larger question — How long does it take before to get a job in Data Science ?
We request and we set the expectation with the learner that please have a plan of focusing yourself for the next 6 to 9 months to be able to make this sustainable transition in Data Science because you don’t want to be taking unnecessary pressure burdening yourself.
Short answer 6 to 8 months.
If you are 2020-2021 pass out, just out of college and you want to make a career in Data Science.
Number one, most likely, your path is going to be through an internship, you start off as an intern, and then you will be able to convert that opportunity as a full time role.
Number two, you should not be looking at your compensation in the first Data Science role that you get as something of a major springboard in your life.
Number three, you should definitely value the mentors, the projects, the team, the domain, the kind of work that you get to do more than anything else in the first year itself.
Having said all this, let me just put a number, how much salary will you get as a fresher?
If you’re starting off as an intern, the internship can pay anywhere between INR 2 -3 lakhs a year on a higher side, that’s an internship.
If you’re starting off as a full time role in a company services or product as a fresher, you should expect a compensation of anywhere between INR 4 lakhs to 8 lakhs a year on the higher side and that’s a very, very, very good start but remember three points that I spoke about.
It is all about knowledge. It is all about the kind of people that you work with more than the first compensation that you get fresh out of college.
Whether you are from a non-technical background, you are from Bcom, BA, BMS, or you’ve done completely different psychology, sports, anything.
The most important thing is that you have done your 10+2, or 10+3, in that you’ve covered some very fundamental and basic soft mathematics and some of the other subjects which have a slightly quantitative bent of mind and those are enough and sufficient foundations for a career in Data Science.
Here’s a simple trick.
It is not the maths, but it is the mathematical intuition. It is the applied maths that’s going to be super relevant as you move ahead.
You need to sharpen that intuition and not the underlying skills, derivations, formulas.
So absolutely, if you are from a non-engineering background, nothing to worry about, you can make it big in a Data Science career, and we have enough examples of them and we’re speaking from first-hand data.
The answer to your question is, YES.
It’s a binary YES. You absolutely need a GitHub profile.
Just think about it, you are actually interviewing and you have two candidates in front of you, one of them comes back to you and showcases a GitHub profile, and the other person doesn’t have one.
The one who already has a GitHub profile, has a head start.
A Github profile with minimum, 2 to 4 projects, good ones, very well defended, very well detailed, I think it gives you a very strong start.
So the short answer to your question is YES, absolutely Github profile is necessary and at GreyAtom we also show you how to make it happen, build it over a period of time and make it valuable .
Just use your cell phone in selfie mode. Point and shoot. Ask us any question – We will respond to you and add it right here.