We have heard of so many success stories by far, in regards to the journey of being a data scientist.
What’s the thing that keeps data scientists going even when things get tough? What is so intriguing about the data industry that made so many people venture into it even though they are from various backgrounds; and what makes them decide to stay in it despite the challenges?
Today, we have Ghazal Ghalebandi, data scientist from Job Street, to share with us her journey in becoming a data scientist, how she persevered through the hardship when she first started, and what she did not expect in her career in this industry.
Can you share with us a little bit more about your background and what led you into data science?
My major was in computer science and data communication, and I decided to venture into social network analysis during my Masters studies. That was when I was introduced to R language programming, and for the first time ever, I really enjoyed learning programming.
It was through learning R, I was introduced to data science.
I became really interested in the field and went on to learn other languages like Python, and even going into various machine learning concepts, algorithms etc. I learned from online courses, attended workshops, joined meetups related to data science; just to get as much exposure as I can to this subject matter.
It was also through one of the meetups that I found my job in JobStreet. I first joined as a junior data scientist, and at that time, I was still juggling my PhD studies in interdisciplinary research in information science, psychology, and education (which had nothing to do with data science).
Because I had to juggle both, it was really quite challenging. But I was truly interested in learning about data science and how it was able to solve problems in such a systematic way. That was what kept me motivated to keep learning and becoming a data scientist today.
Watch: Do you tend to use Linkedin after every conference?
Besides your strong interest in data science, what made you keep going despite the challenge?
A big factor was the company I joined. I was very lucky because the entire team and the culture of the company is very supportive.
Skewed towards a learning organization, they allow their employees to learn and grow, plus provided the necessary technology and tools that gave us an opportunity to explore.
It was also a safe space to make mistakes and learn from them. The senior employees would always be open to letting you go through some trial and error, as long as you are able to learn and progress from there.
I think it is incredibly important to have mentors to be open for you to explore and learn because it’s then, you would learn the most, and find accomplishment in seeing the progress you have achieved throughout the process.
This sense of accomplishment would also further motivate you to power through even when it gets really challenging.
Now to the most debated question in programming… R or Python?
Based on my own research, I see that more companies are using Python as the primary language. So to a certain extent, Python is a good basic skill to have, but learning other languages on top of it is always helpful.
It also depends on the kind of work you are looking to do. For example, if you are looking to do more statistical and analysis work, then R could be more helpful. But for the most part of the data science project life cycle, Python would be sufficient enough.
But I genuinely think it doesn’t matter what language you learn or use, as long as you can explore, experiment, build your model and produce the necessary result.
Just as advice, it’s always good to be open to learning more than one language, and you can always further explore into whichever you are more interested in and go from there.
Should beginners learn programming first, or take a step back to learn data science fundamentals?
Well, programming is one of the fundamentals of data science, together with communication and business skills, computer science, statistics, etc.
You cannot just do one of it, but neither can you do all of it at once. So think of programming as one of your tools in understanding data science, as it can help you to build, model and further explore the data you have. Then from there, you can look into what kind of problems you are looking to solve and expand your skills accordingly.
One of the mistakes I made while learning was that I was greedy and wanted to learn everything at once. That made me only mastered very surface-level skills of each of the fundamentals. I would definitely recommend people to look into one component, slowly take time to learn and understand it well before moving on to the next one.
But if it does get messy during your learning journey, remember that it is very normal.
There are so many different dimensions you can venture into, and so many available tools out there you can explore. However, as you learn more about each part, you will be able to figure out which area you are most interested in, and you can slowly narrow down your specialty.
Based on your experience, how should organizations new to data science, start?
First, the company needs to look at the problems they are trying to solve together with exploring the necessary resources, such as the type of data and technology they will require to handle the data.
In terms of the size of the team to build, a small team of 3 should be sufficient in general, which consists of a data scientist, a senior data scientist and maybe one more experienced principal data scientist.
With a fairly experienced team lead, you would have someone to coach your team to kickstart the project and they can also help you get the desired results faster than having to navigate through it on your own.
It’s also important for you to find data scientists that trust you, and truly understand the problem you are trying to solve, and understanding how to convert the business problem into a data science problem. With that, together with the team, they can implement and build a solution that is tailored for your needs.
Watch Ghazal’s presentation on Application Relevancy Metrics in D/M Summit.
What is something you did not expect in your data science career?
This is a very interesting one! I came from an academic background. So previously for me, it has always been about finding a solution, writing a publication and that was it.
But in data science, there are so much more layers to that. Yes, you found the problem, you build your experiment, you build your model… Now what?
The most challenging part of the data science process is that you need to go out and convince people. In this case, you need to have business acumen as well.
You have to integrate what you have built to a bigger audience, and in some cases, you will need people to buy your products and metrics. This is the part where you need to create justifications on the value your model brings to a business and why they should listen to you. Again, you need to remember they have no technical background, so you will have to translate it into something that makes sense to them.
When you first start out in data science, this is not something you may think is important, but once you get to a certain stage this is quite a challenge and it is also a crucial part of being a data scientist.
Hence, it is so important for people who are looking to venture into this career to build their communication and storytelling skills.
There you have it, a sneak peek into the journey of Ghazal as she navigated through learning data science and becoming the data scientist she is today.
She shares more valuable tips and insights from her years of experience in her session on D/M Summit, so be sure not to miss it. Join us in the first-ever virtual data science and marketing summit in Malaysia.