If you are satisfied, that’s where the adventure stops.
Dream vs Reality. Journey vs Destination
So, you have thought to accomplish your dream goal. But it’s still an image in your head, right? How will you make it real? How do you clear that blur?
Long term goal also comes with short term setbacks 📈
The journey which life takes you through becomes so much more exciting than the goal that you keep setting higher targets and give a pat on your back per milestone achieved.
Sometimes you don’t even care about the destination and just enjoy the moment 💯
This phase lasted two years for me and now although I feel that I’ve achieved enough, the idea of lifelong learning still prevents me from reaching learning saturation in my day-to-day work.
In this post, I’ll be sharing my journey into the field of data science to data engineering, how I narrowed down to it (data science is still a branch of computer science, right!), the challenges faced as a fresher, and how the IIT Roorkee and Udacity communities together not only helped in tackling them but also proved that it is never late to start afresh, especially when you know to harness the full potential of a vibrant community, both offline and online.
It was May 5 2017, my 4th-semester exams just ended (woohoo n all) and 3 months from then on-campus internship process was about to start. So, all I had were 3 months to master relevant skill-sets (I had a vague idea at that time) and a good experience (say, 3 months 😂) to apply for an on-campus internship of my choice.
Just then another bomb dropped. On May 17 that year, I received F grade in one academic course and went on a kind of downward spiral. That hit me at the core and that’s when I realized, “Well, what to do now?”. Yeah, just a simple setback. It matters to me because I’ve never received such an evaluation in the last 10 years.
I lost hope that I would even get an on-campus internship in tech field based on my experience (I secured an off-campus one eventually)! I am from Engineering Physics (B. Tech) background and that too with less than 70% grades. So, there was not just one reason to panic. College folks in my country know that.
For the time being, I started exploring trending technologies which were in demand at that time. Came around some hot topics like blockchain, deep learning, artificial intelligence, data analytics, Dapps, DevOps, business intelligence, machine learning, digital marketing (this caught my attention), etc.
I started learning a few things like Digital Marketing (Internshala Virtual Training Centre), Advanced Excel (same source) and Data Analytics (same). I started applying for internships in all these areas. After receiving loads of rejection, I narrowed down my area of interest to data analytics (long story cut short 😅) after seeing my interest and deciding long term focused career objectives.
Then I heard about Coursera and Udemy from my colleagues. Learned the math concepts behind machine learning algorithms from Coursera and went on a kind of audit mode after 3 weeks, because the assignments were in Matlab and I didn’t know Matlab that well to code and test my results. I wanted to apply it side by side in a comfortable scripting language, preferably Python or R. I don’t know if Coursera modified the curriculum now.
Eventually, I didn’t complete (on Coursera) and tried Udemy. The problem with Udemy was that the solutions were readily available in the next video. So, there was no driving force to attempt the assignments on my own (I did Data Science and Machine Learning Bootcamp with R, can’t say about other courses). Anyhow, I finished R programming on Udemy.
Forget Python vs R. Its Python & R.
Use of language depends on your comfort, community support and your organization's technology stack overall to ensure collaboration across different teams.
Python is a multi-paradigm language with applications in various fields (this site alone shows eight applications). So, when I finished revising all concepts in Python (from Codecademy), I was confused about which track to follow. With R, I had a laser-focused path to learn statistical computing and stick to data analytics.
Later I learned Python and SQL as requirements changed.
Other Online Platforms
As you can see from my previous post that Udacity was not the only thing I tried. Before that, I tried other online platforms too. The problem which I found on other free platforms initially was:
- transparent feedback
- engaging mentorship
- performance evaluation system
- personalized support
- turnaround time for queries
At that time, I already knew about Udacity but was reluctant to spend too much money on their courses. As soon as I heard about their one-week full refund plan, I thought of giving Data Analyst Nanodegree a shot.
One week into it, I found the first project daunting, like hitting an impasse after every couple of TODOs. Anyhow, I completed the first project within one week and the positive review highlighted my efforts and encouraged me to learn further.
After completing the first two projects, I really absorbed the curriculum and grew with it.
I utilized the full facility of mentorship support via chat, 1:1 video calls, peer support from Data Analyst Nanodegree students.
The projects became more technical with time and difficult to work with. That was the time I needed to relax. My mid-semester exams were also approaching, so I took a 1-week break from Udacity and came back after finishing my exams. And with a fresh mind, I just revisited the basic concepts and started working on the remaining projects.
Side-by-side, they also have a Career Portal where you can learn networking skills and online branding. After completing 4 projects, I thought of showcasing them on my profile. So I started optimizing my GitHub and LinkedIn profiles for “data analyst” position. And yes they have project reviews too for these profiles too!
Career Services comes with resume review, cover letter, LinkedIn & GitHub reviews (and some reviews I still can’t remember because they were not relevant to my career path). My first review was terrible for all platforms. So, I submitted my second review for all of them after nearly 5 months (those green squares on GitHub 💚).
I can say its a one-stop package for your online professional portfolio optimization.
While starting DAND, I found the lecture parts easy because of my prior experience. So, to engage myself, I enrolled in ML Basics ND. I completed the Machine Learning Basics ND (2 months program) in 10 days because of my previous experience in ML with R (thanks to Udemy and Coursera).
But then I took a risk which I would never recommend anyone to do ❗️
I enrolled in Machine Learning Advanced Nanodegree 😅
Doing two Nanodegrees together (and that also with one of them being of non-beginner level difficulty) can be nerve-wracking sometimes.
Multitasking is not always good
If you can work with 🎧, you’re not really a multi-tasker 😂
Both, ML Advanced and DAND Term 2 were math-heavy so I really cut-off all leisure time, sleep time, and social interactions to complete them. As soon as I completed DAND in January, I started applying for internships and received callbacks (after loads of rejections 😅 ). Rejections at the start, but finally, an IIT Roorkee alumnus offered me a position of Data Science Intern at their startup.
Then what, I considered it and took 24 other IIT Roorkee students with me for interviews. Out of which 7 were selected (excluding me, I worked as an organizer for that event).
Outside of Nanodegree Curriculum
Diversity is good for exploration and whatnot.
The projects in the curriculum are quality ones, but a real-life project outside the curriculum really shows what you have learned from it. So, I picked up an on-campus project side-by-side on a real-life problem faced in India at that time: train accidents. You can learn more about it here.
The main challenge was data munging. I decided to complete the entire project in R and took around 6 months to complete it. These 6 months were very hectic. I networked on LinkedIn, did some cold calling to recruiters and improved slowly on how to optimize my LinkedIn & GitHub.
People asked this question many times. How did I manage multiple Nanodegrees financially?
A simple answer to this question would be just perfect timing.
In Data Analyst Nanodegree and Machine Learning Basic Nanodegree, I enrolled at early-bird prices.
While I was in the mid of Data Analyst ND, Udacity India hosted a Data Challenge, cracking which I received 100% refund for Data Analyst ND.
Around January 2018, Udacity announced Google India Challenge Scholarship. It was basically meant for anyone who is motivated to up-skilling their Android or Web Development skills. I got selected for phase 1 and enrolled in their challenge course: Android Basics ND. I enrolled because I just wanted to try my hands on Android development as a side gig. Turns out I am not into Android development that much. As a part of this scholarship, Udacity offered discount on successive Nanodegrees after I couldn’t qualify for phase 2 of Google India Challenge Scholarship.
They still wanted me to feel motivated and keep learning.
Based on your some basic level Nanodegree completion, they enrol you directly in term 2 of their advanced courses. Don’t know about now since they switched to a monthly subscription based enrolment.
So, yeah if you have money problems and still you want to pursue a paid course which you really like, look out for scholarships or offers for the same.
You have to invest either time or money. One resource saves another.
What to experiment is entirely up to you.
Udacity announces scholarships which help you enter into their career ready Nanodegrees for free (after shortlisting from scholarship). Students passing the PyTorch Scholarship Challenge from Facebook are currently doing Deep Learning Nanodegree for free. See his reaction.
Five Nanodegrees and still going…
I don’t know where lifelong learning will take me.
On June 6 2018, when I was having lunch and scrolling my Flipboard, I saw the program I had been waiting for so long. Udacity launched Data Scientist Nanodegree (DSND) program. Then what, with the offer from Google India Challenge Scholarship in hand, I entered DSND at a discounted price and continued my journey in data science with Udacity.
Just then Udacity launched AI for Trading Nanodegree (AITND). I meant what!!! Finance + Data = ️🔥. So enrolled in AITND. I will do its term 2 later. But as of now, I am interested in exploring the data engineering part in more detail after completing the Spark for Big Data Capstone Project in DSND.
As a whole, I really liked the content of Data Analyst Nanodegree, though some parts like hypothesis testing and confidence interval required multiple revisions.
Currently, I am doing Data Engineering Nanodegree along with my job and hope they launch more programs related to this field by the end of my program 🤞🏼.
IIT Roorkee Community
I enjoyed my first two years in college (mother of laziness) like I am on vacation. I was not actively involved in any cultural or technical groups. But when I started my journey in data, I received great support from all of them, especially my dear friend Dhruv Bhanushali (not in above) from IMG, IIT Roorkee. He was kind of an offline mentor during the initial stages. I disturbed him a lot. I still do.
IITR’s central library is where productivity peaks. Two years ago, I experimented staying there for seven days straight. No conditions, just a seven-day commitment. And then, people knew where to find me.
Formal education will enable you to earn a stable income. Self-education will enable you to amass a fortune — Siraj Raval
College education combined with online learning instills tremendous knowledge one can possibly need. One provides stability and security, and another exposes you to industry trends.
Finally, I got a chance to share my experience with Udacity Student Success Team in this interview. The interview was fun and they created a nice blog out of it. Hope you like it.