Original article can be found here (source): Artificial Intelligence on Medium
1. Should you become a Data Scientist?
The very first question you should ask, why a Data Scientist?
The hype around the topic of data science is very real but that should not be the only reason you are coming into this field. The fact is, becoming a Data Scientist, especially from scratch, is no easy feat. A Data Scientist requires a certain knack for looking at data and it involves lots of experimentation, research, critical and analytical thinking. The media often portrait data science as the cool kids next door, working on amazing technology such as machine learning or artificial intelligence. But behind every buzz words are tons of programming, mathematically theories and experimentation, most of it don’t rhyme well with cool.
However, if you have done your fair share of research and decided that Data Scientist is right for you, here is the good news.
The best time to get into data science is NOW.
Other than the fact that;
- Linkedin reported that Artificial Intelligence Specialist is the #1 Emerging job in 2020,
- World Economic Forum predicted that Data science and related roles created a net 58 million increase in jobs in 2022, and
- Glassdoor revealed that the median base salary of a Data Scientist is $107,801
Data Science has also matured substantially in recent years with its potentials greatly recognized by most industries. Areas like healthcare, finance, agriculture, retail, defense have already seen their operation impacted by AI and there is only more to come. This means that equipping yourself with data science skills is valuable to your employers, no matter the industry.
Furthermore, learning data science has never been easier! Just look at all our posts on data science courses and books available at your fingertips. The immense interest in data science has led to an explosion in the amount of data science resources, benefiting both beginners and practitioners in the field. Videos lecture, blog articles, and e-books, a wide variation of teaching materials are publicly available for learners. To add on, offline in-person lessons are also in-demand right now and we will introduce them in the next few sections. Keep on reading.
So should you become a Data Scientist? If you believe in the potential of Data Science, if you are motivated and driven, if you are excited about changes, I will say why not?
2. Your desired role in Data Science
Even though we generalize everything as Data Scientists, understand that there are multiple roles in Data Science.
- Data Analyst
- Data Scientist
- Machine Learning Specialist
- Data Engineer
- Business Analyst
- AI Research Scientist, and many more
Specifics of each role are beyond the scope of this guide but this blog post summarized it nicely.
Why is this important? The role you chose decides the skills you need to hone. A Business Analyst concerns more about their products and business knowledge while a Machine Learning Specialist focuses on their technical skills. Having an ideal role early in your journey helps to plan your learning, projects and job hunt later on. Other than that, you can even go further by narrowing down the specific industry for the role you want. One piece of advice is to leverage the prior experience that you may have. If you are knowledgeable or passionate in certain fields, find data science-related roles in these industries that integrate your experience and data analytics skills. This way, you will definitely be a valuable and in-demand asset in the industry.
Do some soul-searching at this stage and find the role you really want. Although this guide focuses primarily on Data Scientist, many of it still applies to any other roles.
3. Qualification Check
Check what now?
This is the time to assess your past qualification and see where need to start your data science learning journey. Doesn’t matter if it is a completely fresh start, but it’s important to know where.
Use this infographic to find out where should you start. Basically, you should try to check off the few fundamental skills needed to become a data scientist.
Many of you will be doubting yourself and question if your qualifications are sufficient to build a career in data science. I can and I will assure you right now that you definitely can. Apart from the more academic work such as an AI Research Scientist, you certainly do not need a Ph.D. or even a higher degree to excel in data science.
Proving our point, Elon Musk recently tweets this when asked about the invitation to an AI hackathon hosted by him.
Elon Musk has spoken, so drop your self-doubt and get started.
4. Level up your Data Science Skills
Once you figured where to start, its time to get into action. There are typically 3 ways you can go about learning Data Science.
- Graduate Degree (Master or Ph.D.)
- Massive Open Online Courses (MOOCs)
These are rank according to the ability to personalize your learning with 1 as the least customizable to 3 the most flexible. There is no one perfect path, just the one most suitable for you.
Bootcamps are basically intensive on-campus training that ranges from a few weeks to a few months. They aim to cover as much content as possible in the short duration so you would expect a very steep learning curve. Time is also a factor as these often require massive time commitment during the course. As such, this option is only recommended for full-time learners who are able to commit to their schedules.
In addition, this is also the least personalized option on this list. Mainly because most bootcamps assumed zero knowledge and attempt to teach everything from scratch, so there is no need to customize your experience as everyone starts equal.
One plus point of bootcamps though is the opportunity to connect with like-minded individuals. Your fellow learners should be as motivated as you are, who wanted a change or advance in their career and all of you are working towards a common goal. These are the best candidates to network with as you can be sure they will be active in the data science field.
Graduate Degrees usually span from 1–2 years for a Master’s degree to 3–4 years for a Ph.D. These are specialized post-graduate programs with validated curriculum to teach the specialization you chose in-depth. Some examples are the MIT Master of Business Analytics and Master of Science in Data Science from Columbia University. Furthermore, these programs provide some degree of personalization where you can choose the specialization you want. You can take modules that interest you and personalize your learning. However, stating the obvious here, this is also the most expensive option in this list and you have to take that into your consideration.
Enrolling in a graduate degree program is often a huge commitment, so make sure you have done your research.
Massive Open Online Courses (MOOCs)
MOOCs have come a long way since it first started. Recognized universities are releasing some of their modules as MOOCs and even major tech companies like Google have come up with their own online courses. Gone are the times where MOOCs were being frowned upon and companies are becoming more receptive of self-learners. Using the right resources and the right learning path, anyone can build their arsenals of data science skills using online courses.
In addition, there has been a shift from on-campus teaching to online-style delivery. As such, for people who wanted the flexibility of online courses but the credibility of a university, edX or Coursera have partnered with Universities to provide their Master’s programs on these platforms. This has brought MOOCs to a new height and changed the definition of what constitutes an online course.
5. Work on Projects
While courses tell employers you know data science, projects show them your ability as Data Scientist. Of cause, doing projects is not only for your resume but it also builds up your technical skills, maybe more so than any course could.
Kaggle, the best data science competition platform. Even though Kaggle is known for its competition, working on Kaggle can be a project on itself. Competition datasets are real data by companies with the purpose of using the strength of the community to solve their business problem. During the competition, you will experience the process of data collection -> data processing -> modeling -> evaluation -> and optimization, similar to a self-initiated project.
If you are new to Kaggle, competitions are available in few categories. For starter, there are playground or knowledge competitions that test your fundamentals while featured competitions are available for Data Scientist to compete and win prizes. If possible, work on active featured competitions as ranking yourself against the leader board is a good experience and a great addition to your resume.
However, the major drawback is that datasets from Kaggle are just too clean. Basic pre-processing of data was already administered by companies before hosting the competitions in Kaggle, reducing much of your workload. If you have not heard, Data Scientists typically spend upwards of 70% of their time processing data while only 30% of the time modelling. So you can see that being spoonfed in Kaggle might not be the best for you.
Hence, I also recommend you to go out there and work on some self-initiated projects. It does not have to be game-changing or life savings. Just novel and innovative will do the trick. If you know what industry you wish to work in, find projects related to that industry. Find your dataset or scrap a website, go through the process of experimentation and experience what is it like to be a Data Scientist. There is no better advice than having your own beloved pet projects which you can proudly present to the world (and recruiters).
6. Build up your Online Presence
What do you mean by online presence? Do I need to be an influencer to be a Data Scientist?
Not quite, but close. Having an online presence has become increasingly important in the tech world, including Data Science. For self-taught learners like us, an online presence helps to validate our work and qualification in data science. This validation comes in the form of the followers, comments, and peer-reviews you have in the online space. But this is not any social media followers, I mean people who read and find your articles interesting, people who share the same thoughts or issues as you in the field, or even people who get inspired by your ideas and use your projects.
Yes, the easiest way to build up your online presence is to write your thoughts and share your work.
Medium is an online publishing platform where writers can write and share their articles at no cost at all. You do not need a domain name or a hosting server to post your content online. Just sign up an account, build your profile and write away. Medium has also become the preferred platform for publishing data science articles making it the best choice for you to start.
So what to write? It can be anything or everything. The main goal is to pen down your thoughts, your journey and appeal to people like you.
Some blog ideas;
- Share your Data Science curriculum with others
- Reviews courses you took
- Short write-up of your Data Science projects/competitions
- Thoughts on Data science-related matter
So what do you think I am doing here?
Oh yes, Github.
You do not truly belong to the technology sector if you don’t know GitHub. GitHub is a development platform where users upload their open-source codes in the form of a repository (repo) to manage or share their projects. Furthermore, It has great build-in version control functions that benefit both software engineers and data scientists and allow collaboratory works between professionals. Hence, starting a GitHub account goes a long way to getting noticed in the tech space. Some potential employers might even ask for your GitHub to assess your projects, codes and technical competency!
So what are some of the things you can do with GitHub?
- Upload projects
- Fork interesting repos
- commit to other repos
7. LinkedIn and Networking
Never underestimate the power of networking. When walking into unknown territory, having offered a helping hand is rarely a bad thing.
This is where LinkedIn excels as a social platform. LinkedIn allows users to build their professional profiles and connect people with similar backgrounds. You can find people in similar space, send them a personalized message and build meaningful connections all in a single platform.
Here are 4 tips to help you optimize your LinkedIn profile.
1) Add a Professional Profile Picture
Firstly, when logged on to LinkedIn, click on ‘view profile‘ to edit your profile.
A profile picture and a cover image are used to represent yourself on LinkedIn. This is the first thing people see before reading your profile and a good first impression counts.
2) Fill up your profile sections
Your profile is segregated into sections where each section tells something about you. The most important sections to tell others all about yourself are your experience, education, certifications, and projects. To start, clicks on the ‘Add profile section‘ and add these sections individually to your profile.
You can find ‘Experience‘, ‘Education‘ and Certifications under Background while ‘Projects‘ can be found under Accomplishments. Flaunt your achievements, this is your profile and be proud of what you have achieved. Nothing is too insignificant, give yourself the credit you deserved. With that said, your profile should remain factual. Honesty is the best policy.
3) Personalize your Intro
Next, personalize your intro by pressing the pencil symbol under your cover image. Change your headline to reflect your current status. It can be your current role in the company, current education or even general statement about yourself.
- Data Analyst in Self Learn Data Science
- Computer Science Major at Stanford University
- Data Science Enthusiast connecting with Like-minded Individuals
Anything that describes you really. The headline will appear alongside your profile picture so here is also where your first impression goes. Make it count.
4) Let recruiters find you
After completing your profile, its time for recruiters to find you. The last step is to let recruiters know that you are open to opportunities and the job you prefer. Head back to your profile page, click on the ‘Add profile section‘ and find ‘Looking for job opportunities‘ under Intro.
Once set up, it will appear on your profile but only visible to recruiters.
Now you are done. Pull out your contacts or head to suggested networks and start connecting. Happy linkeding.
LinkedIn is hardly the only way to network. In fact, connecting with others face-to-face leaves a more impactful impression and certainly more meaningful connections. Join your local data science community, attend a meet-up or go to a conference. Even if you are using LinkedIn, do not miss out these channels of networking. All are great avenues to meet people, gather ideas, and who knows, you might find your next boss in the mix.
8. Build your Resume
Finally, you are ready to get a job in Data Science. Or is it?
There is just one last hurdle — getting your resume noticed by employers. Recruiters and employers get hundreds of resumes per job opening, so how do you make your resume stand out from the rest and not get filtered to its demise.
Other than your name, make sure to add your social and online profiles. This is to allow recruiters to find you easily if they wanted to.
2) Projects / Experience
if you do not have any data science-related experience, we recommend prioritizing your data science projects before experience. This way, it will be kept relevant to the job scope for whoever reading it.
List down your top 3 projects starting from the most impactful. Try to have some variations in your projects to show a wide range of experience.
For each project;
- Make sure the title is self-explanatory (E.g. Using Twitter’s tweet sentiment to predict prices of S&P 500)
- Describe the project: Start with a strong action verb -> Tools you used -> results you achieved
- Keep it short, concise but show your competency objectively
3) Experience / Projects
The Experience section will come next, list your 2–3 most relevant and recent working experience.
Here is where your blogging experience can come in handy if you do not have any tech experience. It might not seem much but it will definitely show your enthusiasm and motivation in the area of data science.
As for other experiences, keep it relevant to the employers. Even if your experience is not related to data science, rephrase it to advertise your soft skills such as communication, leadership or time management.
Be strategic and relevant in this section. Although you can pretty much write anything here, keep it to the most important few especially in the ‘Certification & Course‘ section. Unless these certifications are widely recognized, most employers do not care what MOOCs you have taken. Only state those that are popular or from a highly credible university.
As for technical skills, do not spam all skills you think you know. Be strategic. Use the company’s Job Description (JD) as a reference and write those that are required for the role you are applying for.
Try to keep your resume within one page, the most two. Concise is key and personalize each resume to the role you are applying before you hit the send button.
9. Get Hired
Now you are ready to be hired as a data scientist, go ahead send your resume and hope for the best. As a self-taught data scientist, my advice to those walking the same path is to persevere. Data science is never easy and so is getting hired as a self-taught data scientist. Assess your ability and do not be afraid to start lower.
Don’t wait, be motivated, get working and join us in this exciting career of a Data Scientist.