Spoiler: Kaggle wasn’t considered one of them

Over the past 3 years, I actually have modified careers from working in Project Management/Data Evaluation to becoming a Data Scientist.

In this text, I’m going to speak about 11 things that helped me make this transition and get my first Data Scientist job.

There’s no option to sugarcoat it: getting a job in data science/ML/AI is difficult, and it would require loads of work. But should you’re willing to place in the hassle, this text will assist you be more strategic about the way you spend your time and energy. By sharing my very own personal experiences, I need to attempt to assist you cut through the noise and the marketing fluff that surrounds the industry. Because data science is such a hot industry, there are loads of people attempting to sell you stuff (courses, equipment, data, certifications, etc.), and in my experience this makes it really difficult to know which advice to trust, and which to disregard.

In this text, I can promise you that I’m not attempting to sell you anything. My aim is just to provide an honest account of my personal experience getting began in the sector, within the hope that it helps you in your journey.

With that said, let’s get into it.

The very first thing that helped me was coding on daily basis. Well, perhaps not every day; there’s more to life than data science. But should you’re fascinated about entering into the sector, there’s no higher option to prepare than coding recurrently.

Why? Well, vulnerable to stating the apparent, because that’s what you’ll be doing most days as a Data Scientist.

The added bonus of that is that it’s a extremely good option to test the water and see should you actually wish to change into a Data Scientist. You is likely to be drawn to the profession by the large salaries or the hype around technologies like ChatGPT, and while those aren’t bad reasons to get involved in AI, it’s also really vital to think about whether you’d actually benefit from the day-to-day work of being a Data Scientist. By spending just 20 minutes every day coding, you’ll quickly get a way for whether you benefit from the sort of work that Data Scientists and ML Engineers do.

Which coding language(s) must you learn? You’ll get different opinions about this, but I’d recommend starting with Python. Spend a while getting used to the Python libraries pandas, numpy and matplotlib, for instance by completing the Open University’s free online course Learn to Code for Data Evaluation. Then, when you’re beginning to get the hang of those, move onto the fun stuff: machine learning. Start with scikit-learn after which, once you begin losing interest of model.fit() and model.predict(), have a go at some deep learning with PyTorch or TensorFlow.

Alongside this, try learning SQL, which is utilized by Data Scientists for data collection and preprocessing. Together, Python and SQL will form the bread and butter of your day by day work as a Data Scientist, so that they’re a wonderful place to start out.

The following thing that helped me was incorporating data science into my existing roles. My advice to you can be the identical: don’t wait until you’re “officially” a Data Scientist before you begin considering like one. As an alternative, change into a ‘Citizen Data Scientist’ by findings to incorporate data evaluation in your on a regular basis work.

For instance, you possibly can volunteer to tackle a few of your current team’s reporting tasks, or try performing some evaluation that helps refine the team’s strategy. Once I was in my previous (non-Data Science) team, I became “the info guy” by offering to assist individuals with their analyses and making some interesting reports and data visualisations. In fact you’ve gotten to balance this along with your day job’s responsibilities, but should you can discover a option to incorporate data, that is a fantastic option to add value to your current team while also constructing your individual skills and experiences.

When you work in a giant company which already has a Data Science team, a straightforward option to change into a Citizen Data Scientist is by reaching out to someone in that team, explaining your ambitions and asking whether there are any projects that you possibly can support with. Chances are high that they’ll be greater than completely satisfied to must your support, because it’s a little bit of a win-win situation for them, too.

This was a game-changer for me. In my previous article, I wrote about how making a portfolio helped me gain traction with recruiters and ultimately land my first Data Scientist job. It also helped with motivation since it gave me a tangible “thing” I could continually work on and improve.

There aren’t any strict rules on what to place in your portfolio. You may fill your portfolio with projects from online courses, side projects, whatever really. The more interesting/unique the higher, but equally don’t kill yourself attempting to do something crazy fancy. The portfolio isn’t going to get you the job; it’s more about starting a conversation and getting your foot within the door.

When you’re stuck for ideas, try searching through some projects on Kaggle. Personally I never used Kaggle as I used to be in a position to fill my portfolio with coursework projects from the varied data science and ML courses I used to be taking, but Kaggle is unquestionably a fantastic resource should you need ideas.

You is likely to be considering: why does networking matter? Can’t I just learn online?

Once upon a time, I may need agreed with you. However the thing is: loads of the recommendation you read online won’t be suitable for you, either since it’s (1) old-fashioned, (2) not relevant in your geographical location or industry, or (3) making assumptions in regards to the reader which don’t apply to you. The world of AI changes fast, and you could speak to local Data Scientists to get an up-to-date, personalised view of the industry. I do know this could be daunting, nevertheless it’s super vital and can assist you enormously. Put it this manner: you don’t want the primary Data Scientist you meet to be the interviewer on your first job!

If the concept of networking freaks you out, don’t worry, it did for me, too. Nevertheless it’s actually loads simpler than you would possibly think. Start by following some Data Scientists and AI publications on social media. This is an excellent option to keep your finger on the heart beat with what’s happening within the industry. By way of publications and platforms, I’d particularly recommend following Towards Data Science and HuggingFace. For influencers, you may start with people like Andrew Ng, Chris Albon and Jay Alammar.

Next, should you work in an organization that has a Data Science team, reach out to some of individuals within the team and ask in the event that they fancy a coffee to talk about their experiences. If there’s nobody suitable at your organization, try reaching out to Data Scientists at other corporations on LinkedIn. You may must try a number of before you get any responses; persons are very busy! But persevere and eventually you’ll find someone willing.

Also, don’t worry about “getting it fallacious”. Do not forget that every great data scientist and AI pioneer started off in just your position: a beginner. People generally have natural empathy for those in situations they’ve been in themselves, and my bet is that you simply’ll be surprised by what number of persons are willing to provide a helping hand.

I’d be lying if I pretended this didn’t play a giant part in my journey. The primary certification I got in data science was the IBM Skilled Data Science certificate. That in turn helped me get into my master’s program, the MSc in Social Data Science at Oxford University, which in turn helped me get my first job.

It’s vital to say that you simply don’t necessarily need a degree in Data Science to get into the sector. In my current Data Science and AI team, I reckon that lower than 50% of the team have degrees in Data Science or ML. A more typical path to follow (especially should you have already got one other degree) is to finish some online courses and bootcamps and “top up” any missing skills. If you need to know more about why I made a decision a master’s degree was right for me, take a look at this text where I shared some thoughts on this.

When you don’t have any prior industrial experience in data science, it might probably be difficult to persuade employers to take you on. That’s where internships can really help. In my case, I did an off-the-cuff unpaid internship during my master’s programme, after which a paid internship for 4 months immediately afterwards. Each of this stuff helped me persuade future employers that I had what it takes, and likewise helped me learn some practical data science skills that are difficult to learn through online courses (e.g., git/GitHub and GCP/AWS).

To seek out internships, you’ve got two options. First, you may apply to internships posted on public jobs boards liked LinkedIn, Indeed or Otta (my personal favourite). The issue with this strategy, nonetheless, is that everybody else is pursuing the identical strategy, and it might probably be really hard to get an internship when there are tons of (and even hundreds) of others applying for it.

The second option is to succeed in out to smaller, less famous corporations directly. This has at all times worked thoroughly for me. Write a friendly email explaining what you’re on the lookout for and what you may offer, forged a large net, and also you’ll find something.

Also, I need to acknowledge that taking an unpaid internship isn’t going to be an option for everybody. If that is you, don’t feel disheartened: there are many other ways to get the required industrial experience. For a start, you may take a look at these free virtual experiences compiled by Yash Gupta:

Don’t just apply to the large names or to places that everybody else is applying. Take the time to take into consideration what you’re really on the lookout for, and how much environment you need to work in. For me, after a number of reflection, I felt strongly convicted that the Media sector was where I desired to be, so I only applied to corporations in that sector.

When you’ve decided where to use and submitted your applications, study the job spec and learn all the things you may in regards to the interview process. You don’t wish to be caught out with any unexpected surprises!

Even should you haven’t worked in Data Science before, I can guarantee that your experience will probably be relevant: you only have to seek out a way of communicating that.

Back yourself.

For me, this meant emphasising the communication and storytelling skills I had learned through making umpteen presentations, and the stakeholder management skills I’d learned in difficult project meetings. These soft skills are highly relevant in a Data Science context; you only must make it clear how they translate.

Be realistic in your assessment of your skills. You may have tonnes of experience in an adjoining field, nevertheless it doesn’t mean you’ll find a way to make a straight sidestep into data science. Accept that you simply might must take an intermediate step before you land the dream role in Data Science: for instance, you would possibly must take an internship, a less senior position, or a technical Data Analyst role. I’m not saying lower your ambitions — still aim for the sky! But accept that you simply might must take an intermediate step before you get there.

Changing careers could be hard, and it might probably take time. You won’t get there overnight. But with a bit patience and perseverance, you’ll make it.

Take a break day every week. Remember why you’re doing this. Don’t neglect your health or the relationships in your life for the sake of your profession. All that vital stuff.

Truthfully, I’m preaching to myself on this one. But no profession is price sacrificing all of that stuff, not even data science. Within the words of a Middle Eastern sage (Jesus), “What sort of deal is it to get all the things you would like but lose yourself?” (The reply: a reasonably rubbish one).

Thanks for reading — I hope this has been helpful. Be at liberty to let me know what you think that!