UNAPOLOGETIC WOMEN IN DATA SCIENCE SPECIAL EDITION WITH SAMZEE.

Audacious Coder.
7 min readMar 28, 2020

--

“A woman is the full circle. Within her is the power to create, nurture and transform.” ~ Diane Mariechild

Knowing is better than wondering. Waking up is better than sleeping and even the biggest failure, even the worst, beats the hell out of never trying. -Meredith Grey, Greys Anatomy

This is the best quote from one of the women in technology specifically machine learning who am introducing in just a bit.

This article is a bit different from what I have been tackling. This month of March is known for being a women month due to the International Women’s Day that happens on the 8th of March.

During this month of March, I got to read lots of articles about strong women in technology and the impact they have on society and it was just breathtaking. The most important take away I got from all of them is to be persistent and believe in yourself that no challenge can put you down.

So I decided to dedicate this blog to an amazing woman who is in the field of Machine Learning and has inspired me through her passion and commitment in this field of technology.

INTRODUCTION

This is Samantha Zambezi. She is an intelligent Machine Learning Tech Lead working in Market Research specializing in the area of Natural Language Processing. She is also an advocate for women in stem hence STEMINIST especially African women in stem.

BACKGROUND

She started out as an Industrial Engineer working in supply chain analytics as a consultant. She explained,

“ I had a good mentor when I started that always encouraged me to do some courses and be forward-thinking. The work I did for the management consulting firm was restricting and also not very challenging.

So getting into coding and Data Science was an opportunity to work in a more exciting industry and when I started exploring it, I was like "ya this is what I want to do". I then enrolled at the University of Capetown for the first MSC Data Science course and found my first job halfway through.”

This is a very interesting career shift as people tend to think that you need to come from a software development background in order to dive into data science, Sam’s experience shows us that it's not a determining factor.

Q/A SECTION

This are some questions i collected from aspiring data scientists and her take on them.

1.) What is the importance of AB testing in Data Science? Is it a prerequisite?

I would say it is beneficial but not a prerequisite. I haven’t done AB testing myself we have a testing team that is responsible for that. The importance of it is to make sure that whatever changes you are making to an application/algorithm are improvements.

2.) What is the impact of advance cloud technologies like auto Machine Learning? Is data science role likely to be automated soon?

Most of these tools are sold like magic wands. When you try and test them you come to realize, they have some limitations and serve more as assistant tools to Data Scientists that can help reduce time to market for business applications. Plus, Data Scientists primarily need to make decisions about the data before it can be feed into an algorithm, especially for more complex data. Data Scientists are also needed to interpret the results and communicate them to non-technical business/product owners.

Another thought:

Under the assumption that the whole process can be automated, you wouldn’t need Data Scientist. A Data Analyst, Business Analyst can simply be trained to do the job as these types of tools don’t even need you to have any coding knowledge. In this scenario I see the job of a Data Scientist as a pure research one, building models from scratch and trying to solve new problems etc. There will be very few companies that would need this sort of skill. It is important to consider this possibility and shape your career path so that it is not just running sci-kit learn.

3.) You do interviews for Data Science. What key attribute do you look at?

Problem solvers, people that are teachable and show interest in the problems that the company/industry tries to solve. That is just me. I never get the final say but I do assess their technical test. It is easy to see when someone just implemented code. Even if you didn’t achieve the lowest error rate, if there is logic, structure and a methodology that allows you to have some insight into how the person approaches and solves problems that are more key and can differentiate you from other candidates.

4.) What advice would you give yourself as a student?

I only figured out the possibilities of what careers are out there when I finished my 2nd year in Mechanical Engineering. I then later switched to Industrial. I think reaching out to find more mentors, people in the industry when figuring out what to major in would have been useful. However, it is quite difficult to find the right mentors and people especially when you are not in the industry yet.

5.)Which domain are your projects mainly from?

Mainly Natural Language Processing. I work for a market research company where the main goal of the Data Science division is to merge survey data with social media data and other verbatims. Get to work in conversational AI and emotional AI projects. Quite challenging but was a good learning experience. Some things we do are just simple like automating certain time consuming operational business processes.

6.) What does a typical day look like in your world as a Lead Machine Learning Engineer?

It really does vary. Usually starts off with checking logs of models, solutions, etc as well as if they are any bugs and queries from end-users. There is always something. I don’t necessarily get involved in the bug fixing unless it is super urgent. I mostly assist my team and check that they are no blockers. My main tasks are mostly strategic now because we need to find a better way of managing machine learning solutions in production, especially on the scaling and streaming side. So, I am getting involved in testing a lot of technologies and processes that are out there using business use cases.

7.) Which language do you prefer to use and why?

I like both R and python. R is great for shiny, quick statistical analysis and visualizations. Python is more flexible, great for deep learning and testing out the latest things. Python will keep you more relevant in the AI domain.

8.) What are the main challenges that you face in the tech industry generally and as a woman?

Always having the pressure to over-deliver on every task and project. That is quite tough and can really mess you up. There is that voice at the back of my mind that says I cannot fail otherwise I will immediately be labeled as incompetent by my peers. It takes experience and practices to switch that voice off. I have become better at managing this. Another challenge is motivation. You are working in a field where all role models and those with very senior positions do not look like you, there is a lack of diversity in tech and that is not very empowering. It brings doubt like “I think I am wasting my time” it will take me twice the number of years to reach that position because I am not part of the tech bros club. Also, socially at work, you are more isolated.

Do you believe an undergraduate student can become a Lead?

I think anything is possible. However, to be a lead, you need experience not only just working in Data Science but working with people and leading people. Your team members need to know that they can rely on you and be confident that you will provide them with the support they need to execute their tasks effectively. There is another component that is working with business owners, project managers and managing their expectations. I get a lot of “my solution isn’t working the way it did locally…” You need to know what an end to end Machine Learning project looks like, work with architects and spec out cloud services for Data Scientist. This hasn’t been easy for me and I am still learning and doing training especially on the Deep Learning Side. Technology is always changing so you need a way to keep yourself up to date as well as understanding business challenges and changing needs when it comes to Data Science.

ADVICE FOR ASPIRING DATA SCIENTISTS, MACHINE LEARNING ENGINEERS

Machine Learning is a very exciting field that opens up a lot of opportunities. It is also tough and the journey can become frustrating. Remember not to give up and surround yourself with a community that will help foster your drive and encourage you. Always remember why you started in the first place.

You have to love the process and love learning because technology and the world around us is always changing. Data Scientists need to love learning because learning and keeping up is a part of the career.

I am really excited about the journey ahead as I have learned a lot from Sam. Hope you enjoyed this article. You can follow Sam on Instagram @samzee_codes get to interact with her and view her awesome content on Machine Learning.

I would also like to mention an awesome impactful woman Leila Janah who passed on earlier this year. She founded Samasource company that hires people in areas with little opportunity, trains them in AI data input and other computer tasks, and then provides their services to a list of global companies that include Google, Microsoft, and Walmart among many others.

Thank you, Feel free to comment, clap, follow♥

Stay Home and Stay Safe.

--

--

Audacious Coder.
Audacious Coder.

Written by Audacious Coder.

Hyper-active Data Scientist | STEMINIST | Neuroscience Enthusiast | Dancer |Writer of DS/ML articles😄

No responses yet