Original article was published on Artificial Intelligence on Medium
My Road Map to Data Scientist
When you hear Data Scientist, what comes to mind?
A programming guru? A PhD statistician? The list goes on and on.
In this blog, I jump into my experience working on an official data science practicum project and share 5 skills I found important along the way, the 2 leaps out of my comfort zone, and 3 takeaways from my year in data science.
The 5 Skills You’ll Need
I’ll begin by speaking about the hard skills that I found important. The viewpoint I’ve developed is that a data scientist can be anyone in an organization with knowledge in statistics, basic computation skills and practical domain. More importantly however, to excel in this role, one would require a mixture of curiosity, skepticism and comfort when working with messy input.
While learning the technical skills to become a data scientist such as statistics and advanced computing may be obtained through empirical formulas, the soft skills require practice and time on the job.
I like to think of these skills as a road map towards a data scientist driven mindset:
- Embracing the Unknown: The idea of being spoon fed information is non-existent. Often times, your stakeholders’ requirements are unclear and you must assume the role of being the subject matter expert. The quicker you realize this, the more impactful your insights become.
- Curiosity: To tackle ambiguous projects, having a curious nature is vital to seek the right answers. The wow factor is usually a result of unlocking solutions to problems that are unknown.
- Asking the Right Questions: Similar to a consulting role, when a client is looking for a solution, the data scientist is in charge of examining the problem through questioning everything in order to develop the habit of asking the right things.
- Communication: Creating complex predictive models is one part of the project. The strength however is in communicating results in a simple and persuasive way to influence decision makers. For example, in one presentation to our stakeholders, we found it helpful to introduce an interactive approach in story-telling our data for stronger engagement.
- Resilience: This might be one of the most important skills I’ve built on over the past year. When things aren’t going your way, you have to keep trying. We’re lucky to live in a day and age where we have a large community of data practitioners around us. I’ve had countless hair-pulling moments due to a faulty code just to discover someone from across the globe faced a similar problem and someone else was awesome enough to share their solution. I’ve felt a connection with some StackOverflow contributors that I haven’t felt with my own siblings.
2 Leaps Out of My Comfort Zone
1. Transitioning into a start-up
In my previous job, I was one of over 110,000 employees in a company established over 67 years ago. The workflow was quite structured. There were laid out procedures and certain expectations on how to do things. My work, in an Engineering Department, involved following codes and specifications written by the client prior to the project initiation phase. Besides moving to a different working environment and adapting to a different work culture, I also experienced a complete shift in the organizational workflow.
In my Practicum Project, I worked with a young and small tech start-up where things changed quick. Within a few weeks, my team of 7 students became project owners and had to become the subject matter experts. We started off with three main problem statements; however, in a few weeks we shifted projects based on our customers’ priorities. While embracing a mindset of bouncing from one project to another was challenging it first, it was ultimately a great experience and helped me break out of my comfort zone.
2. Learning the Tools to Take on a Data Science Project
While working as an Engineer, I was given the opportunity to lead a mega project overseeing all mechanical designs and procurement activities for a major client. With my prior experience, I felt I could strongly contribute to my data science team by taking a project lead role.
Despite having experience in leading and managing a project, I realized early on that leading this particular project will be an entire different ball game. It was crucial to familiarize myself with the processes involved in a data science project to utilize team productivity and ultimately achieve our project milestones.