How do I transition from software development to data science?

Key Takeaways

are you a software engineer and you want to switch to data science industry i'm going to provide four effective steps that you can follow while still keeping your job and smoothly transition into data science industry i'm myself a software engineer i started learning data science after doing programming you know for 10 or 11 years in fact if you meet any data scientist today their past background will be either software engineer they would be software engineer they learn some math stats machine learning and they become data scientists or their background would be in math and statistics and they learned coding if you're a software engineer the good news is you have achieved a big milestone already which is programming you know programming is kind of the first step to enter the data science industry now you need to learn some additional skills and follow these steps so stay with me i'm just going to go through the whole road map first thing we need to understand is what are different roles offered in data science industry there are three main career path data analyst data scientist and data engineer if you don't know the difference between these three you can watch my video and it will give you a very clear understanding on what is involved in these three a different career track so let's go through step one number one first you have watched the video you have figured out maybe you want to become less a data scientist data analyst is relatively easier and data analysts get guys or girls get less pay compared to data scientists so maybe as a software engineer let's say you are targeting a data scientist career step number one would be while continuing your job let's say you are a web developer or you are a java developer with few years of experience in your company while continuing your job during your free time on weekends or on work days you know after coming back from the work you want to learn the data science skills in parallel and generate data science work in your current job again you need to learn the skills in parallel and generate the data science work in your current job i know what are you thinking you must be thinking but no in my job there is no data science work well hang on let's think about python python is a super easy programming language that is excellent to automate your day-to-day tasks many times you could be a javascript programmer and if you have a small task you might be using javascript or c plus or java to automate that particular task instead of using those languages use python i was using paul scripting language five or six years back but at some point i switched to python and nowadays in my current job if i want to automate anything i always use python and it is important because python is the language you need to learn to enter data science industry you can learn either python or r but if you're confused just learn python python is a very versatile language very very easy to learn i have a complete python tutorial playlist if you watch 15 videos first 15 videos you can learn python in a matter of two weeks it is so much easy simple and then use python in your day-to-day life you are doing a lot of repetitive tasks you know in your software engineering job right now and you need to think little creative and try to automate these tasks using python programming language now you can also build a full-fledged http server in python so let's say you're using node.js to build an http server sometimes companies have requirement on technology stack and you don't have much freedom but you know for non-important work you know for small tasks maybe right instead of writing your http server using javascript all the time try to write it in python use fast api as a web framework or jungle that way you learn python programming language and that server is going to perform as good as node.js many times people have this argument that python is slow well if python is slow why python is being used in machine learning scientific computation where the computations requirements are quite heavy you know it's quite intense and yet people use python because numpy scikit learn all the underlying libraries that you're using for the numeric computations and machine learning or whatever they give you c performance so python is used just to write your business logic and it gets things done really fast so if your colleagues or your boss is saying that python is slow that is just because they are very good in maybe some other diff programming language and they they have this bias that python is slow but it is not really slow you need to do some convincing maybe you will fail first time but just show them that no python can be fast as well and eventually slowly uh people will onboard python as a main programming language in your team so yes don't miss any opportunity to use python in your current job second thing is start using python and pandas for exploratory data analysis if you are a software engineer your software will have some kind of database mysql postgrade or even nosql wherever there is data there is data science let's say you're working on some customers record or some sales transaction you can use python and pandas to generate analytics on top of it and to learn python and pandas very simple just go through my python tutorial playlist and pandas tutorial playlist first 15 videos super easy to understand this is something you can finish in maybe one month and then once you understand the capabilities of python and pandas try using that to perform simple data analytics you know on top of your database now your manager has not asked you to do this but what you're doing is in your free time you are building this analytics and then showing it to your client or your business manager and they are most going to like it if they are sensible people if they don't like it maybe try to find some people of in your company who are working for a data science department see nowadays all most of the companies have data science departments so try to connect with data scientists and data analysts in your company and try to showcase your work to them and say hey i have built this cool analytics you know using python and pandas are you interested in looking at it and i want to learn it would you give me your feedback the person would you know some people might say no but you have to keep on trying and you might find uh one or two helpful people who will be able to mentor you and who will appreciate your work third thing is you can build analytics dashboard on a same database using power bi or tableau so initially we we did explorative data analysis using python and pandas now you can do the same thing and little advanced things in power bi now if you don't know power bi it's again super easy tool to learn my youtube channel has a sales inside power bi project now using this project you know so many i have received so many comments in linkedin that people did this project and they got a job as a data analyst you can learn power bi very easily by following that project structure it's a very practical hands-on project that you can do and once you have done that project you get a sense of you know what bi tools can do sometimes as a software engineer we don't know what bi tool is or what it can do we are not aware about the potential so when you watch this project on my channel i have a couple of other projects as well like personal finance projects so when you do these projects on my channel or maybe you use some e-learning platform whatever you feel comfortable to learn these skills but once you have realized the potential of these bi tools you can apply those skills that you learned there in building something meaningful some meaningful useful dashboards on your database in your current job so on job learning is often the best thing sometimes people will say again my boss is not supportive i mean i built power bi dashboard but he doesn't like it he doesn't show the interest well big deal as i said contact data analyst or data scientist working in your company and try to reach out to them if not them maybe try to network with folks on linkedin you know try to maybe find a job different job like if you if you want to find a different job go in a company that works on data basically there are some companies where you know database size is not big and they're not very data intensive companies for example my company bloomberg is very data intensive it's a financial data analytics company bloomberg is my current employer and even though you work in bloomberg as a software engineer you're learning a lot of data science skills because you're dealing with lot of data so my suggestion is you're a software engineer your company doesn't provide any scope go join another company which is dealing with data a lot as a software engineer only but when you you are surrounded by so much data there is always a scope of data analytics so you can generate power bi tableau dashboard or you can you can do exploratory data analysis using python and pandas now my channel has data analyst roadmap and data scientist roadmap so if you go to youtube search for code basics data analysis roadmap three months or code basics data scientists roadmap six months you'll find two different videos one video is a three month roadmap for a data analyst second video is a six month roadmap for a data scientist these road maps tell you what skills do you need to learn for each of these roles and i have provided all the free resources you don't have to spend a single penny for that using all the free resources you can learn these skills on your own while doing your software job so that was step number one learning skills learning data science skills and applying those skills in your current job to generate data science type of work when i'm talking about skills you know reading books always helps i'm reading this book called practical statistics for data scientist and this book has a lot of useful concepts and i'm loving it so if you're reading the book you get your math and statistics fundamentals clear as well i also have math and statistics playlist for data scientists on my channel where i explain simple things like hypothesis testing or z-score now very very simple language even high school student can understand it easily all you need to do is watch this video spend some time do coding and just practice and have patience you know it might take six months one year but eventually you will get that so step number one again to summarize was learning skills in parallel while doing your software job and trying to apply those skills in your current job so that you can generate some type of data science work step number two is building a solid project portfolio now in step number one if you manage to find some data science work in your current company cool mention those projects on your resume but let's say you didn't get any success okay how do i get projects to work on well again my channel has uh projects for data analyst and data scientist both so for data analysis projects in youtube search code basics data analyst projects and you will find a power bi sales insight and personal finance project if you are looking for data scientist project in youtube search code basics data science projects you will find three projects one is bangalore property price prediction where i build a model which can predict property prices using linear regression and then i built a website i deployed the server to aws it was an end-to-end project you know as if you are executing the end-to-end project in a real company i have another project for doing sports celebrity image classification maybe you can take that project and do a movie you know movie stars classification or you know the image classification for your family pictures so you can use these projects as a reference and take a little different problem statement and customize it and that way you are learning some skills you are having your own unique project and when you mention it on the resume it's not gonna look like you have just copied some project nowadays most of the resume may have like titanic data set project you know the projects are so boring that your resume doesn't get selected uh for the interview so you need to have your own unique project and you can use other projects on the internet as a reference and just customize it to make it kind of unique the next uh important tip is kegel participation i did an interview with tanul singh who was a mechanical engineer and during his mechanical engineering he used to participate in kegel kegel is a basically data science you know coding practice platform and he would participate in kegel consistently and by the time he got his mechanical degree he got a data scientist job why because his kegel performance was pretty good he had a good rank on kegel so kegel participation even helping ngo or helping your relative who wants to build a software so once you build a software once you have a database then comes data science opportunities you know building dashboards building predictive models so when you help someone for free you are getting a real life project experience you can also collaborate with people on discord for example code basics has a discord server the link is in video description below and there is a partner and group finder channel so when you go there you will find many people like you who want to learn so you can make a group you know if you want to go to gym alone versus if you go to gym with two of your friends maybe the second option works out better similarly group study for data science is always effective so find buddies on discord server and make a group and try to do projects together so that was step number two building a project portfolio by all following all these steps you can build a solid project portfolio and have three or four solid projects on your resume step number three is internship so now you learn the skills you have a project portfolio you build a very nice one-page resume okay resume shouldn't be more than one page even if you have 20 years of experience only one page you use this one-page resume to get internship internships in data science industry how do you get internship well network with data folks on linkedin on linkedin there are many data analyst data scientists try to build connection there is a whole technique of how do you build connections with people because many times people send a message on linkedin and they don't get any response well there are ways to connect with people uh you need to always start with helping them first you know you can send a message saying that okay i am a data science enthusiast is there any task i can help you there are so many techniques you know maybe i can make a separate video on that but network with data folks on linkedin that way when their company has any requirement internship etc they can maybe refer you and referrals always work the best so by building a solid connections on linkedin or through in-person networking by attending conferences etc you increase an opportunity to get an internship opportunity once you get an internship you are in a dilemma now because let's say you worked as a software engineer for five years as a net developer for example now you have some salary good handsome salary when you go for internship internship doesn't pay you much so now you have to make a decision that okay you want to take a pay cut for maybe six months or one year however long the internship is and then use on that time period to learn the practical skills you know to go work for a company as an intern solve some real life problems and later on you can get a job job as a data scientist using this internship experience and when you get that job you will make up for your loss so it's a temporary loss that you're taking but eventually you will make up for the loss and i think more than money the most important thing is okay you are you are doing what you like many times i see software engineers who are kind of frustrated they want to do something different they they feel passionate about data science so now you're getting an opportunity to work in a field that you like so that was internship and then step number four is resume building and interviews you have already you know completed internship now internship experience you need to put in a nice way as a first section on the resume again resume should be one page and now you have a very good resume that you can use to find a full-time job you know you go to linkedin jobs there are so many job portals out there connections on linkedin can help you refer to a job so now you're in a position that you can go for a real uh job interview as a data scientist there are so many there's so much material on the internet available you know that can help you even prepare for the interview there are so many even website my friend gaurav san has this this website called interview ready uh there are so many other portals you know who can help you with mock interviews so maybe spend some money and uh get some mock interview practice that way when you go for the real life data scientist interview you are already prepared so this was a four step process that can help you make a transition from software engineer to data scientist now this journey is going to take some time is there is no shortcut you can't do this in two months you will have to spend six months one year based on your learning ability based on your dedication how much time you spend every day i would say spend half an hour every day in weekdays and on weekdays on weekends maybe spend four to five hours you know or to learn these skills because it requires dedicated afford being data scientist is is a demanding thing you know you need to learn math statistics your communication needs to be good domain knowledge there are so many skills that you need to learn and in order to fulfill that demand you have to of course put a lot of afford but if you follow these four steps while doing this york like 926 current software engineering job you can switch to data science for sure there are so many job opportunities coming up in a data science field and you know you can build a bright future now there is one other option that you can take which is you can do dedicated to your master i've seen people who have done let's say you know job as a software engineer for a few years then they quit the job they focus two years they do full-time two-year study in data science and that's fine i know some people can afford doing that some people cannot because they have family and financial commitments but if you are in a position where you can do it then well and good because two years you are dedicatedly studying data science you are going through assignments and even data science degree will help you to an extent that you know it will prove to an employer that this person has spent two years in a dedicated study of data science i have seen people who even do part-time studies some universities offer part-time data science courses where you do your nine-to-six job then you go to university you know maybe eight to ten you do your study maybe two or three times a week i have seen people doing that as well and if that is something you can do well and good i hope this video helped you if you have any question or comment uh you can post in a comment box below thank you