Data science as a discipline — and specific skills in machine learning, analysis, and training algorithms — is in high demand.
This is an area that has grown in popularity over the past decade and is expected to create 11.5 million By 2026, there will be more new jobs in the US alone.
So, what is it like to work as a data scientist, and what do you need to know if you want to start your career there (or transition later in life)?
I ask Navid Ahmed Janvikara senior data scientist from Seattle who works on Amazon’s Fraud and Abuse Prevention team, shares his career journey.
Check out his story and his advice for those interested in pursuing a career in data science.
Spark: Using Machine Learning to Solve Real-World Problems
What motivated you to pursue a career in data science?
Naveed Janvekar: While I was working at Fidelity Investments as a software developer, my interest in machine learning grew.
I have colleagues as analysts who use data to identify trends and it makes me want to explore this area. So I started analyzing my personal financial transactions to generate trends and insights.
This has led to more time being spent on machine learning and how it can be used to simulate repeating patterns to predict future outcomes and use it to solve critical problems at scale.
To gain better expertise in this field, I decided to pursue a master’s degree in information science, specializing in machine learning and analytics.
After graduating, I worked in various analytical positions at various US companies, such as analyst at Nanigans (a Boston-based AdTech startup), business intelligence developer at KPMG, and senior data scientist at Amazon.
The role of artificial intelligence in data security
As a Senior Data Scientist at Amazon, what role does machine learning play in your work?
Naveed Janvekar: Machine learning and data science play a vital role in my work at Amazon.
In the Abuse Prevention team, we use various classification algorithms and deep learning algorithms to detect fraud and abuse on the platform.
Machine Learning Helps Enable Scalability Compared with traditional rule-based and/or heuristic abuse detection, the detection accuracy is higher.
As abuse becomes complex over time, machine learning can help us meet this challenge as we continually retrain models with the latest abuse/patterns.
I have filed a patent for an invention related to the use of machine learning to detect emerging abuses on a platform.
Convey data-driven insights
What unexpected skills or experiences do you find helpful as a data science professional?
Naveed Janvekar: As a data science professional, gaining domain expertise and the skills to communicate insights effectively and simply to business stakeholders has helped me the most.
When I started my data science journey, I paid more attention to the technical details than to being a effective storyteller.
But over the past few years, I’ve realized that being able to convey narratives and insights from data science or machine learning is just as important as implementation machine learning strategy.
Work with algorithms to create change
How should companies adjust their approach in this area?
Naveed Janvekar: In the past, fraud prevention has traditionally been done using business heuristic rules.
If you observe a pattern that occurs frequently over time, you can enter a business rule to flag the same pattern in the future.
However, this is a short-term solution. It cannot keep up with the evolution of fraudulent patterns.
This is where machine learning and artificial intelligence come into play changed the landscape.
Models are now trained using historical data spanning multiple types of fraudulent behaviors, making these models robust and helping algorithms learn complex behaviors that are much more difficult for humans.
Businesses are already using machine learning in fraud detection. They must now focus on aspects such as automatic retraining of models to catch the latest fraud and make models highly accurate.
This helps automate actions based on model output without requiring human auditors to evaluate suspicious entities flagged after the fact.
Working with data and algorithms can be challenging
But what makes it exciting and fun?
Naveed Janvekar: I love data feature engineering, it fuels my creativity.
Based on domain expertise, data scientists can process data in different ways to answer business stakeholder questions, perform exploratory data analysis, discover correlations between variables, and perform feature engineering for better model performance.
Regarding algorithms, I’ve been trying to train different types of algorithms on a training dataset, evaluating them, and digging into why some algorithms are better than others.
This helped me gain a deeper understanding of these algorithms and where they work and under what circumstances.
All of these make my work fun and exciting.
Become a part of the data science community
Would you like to share a useful tip with data science beginners who are interested in the application of data science in marketing and business and may want to improve their skills in the field?
Naveed Janvekar: A useful suggestion is to get involved in research and invention in the field of machine learning, data science field.
Be part of a working group trying to use machine learning to solve problems in your area of interest.
Contribute to their research, get peer feedback, publish papers, and file for patents.
Through these mechanisms, you are actively contributing to the scientific community, continuously learning from your peers, and improve your skills.
Having a data science mentor is also a good idea.
Keep up with SEO trends
How do data scientists stay up-to-date and informed in the SEO world?
Naveed Janvekar: In the SEO world, machine learning helps with query understanding, voice search, and personalization.
Data scientists can explore the application of various state-of-the-art algorithms to SEO use cases to measure the efficacy of new algorithms.
Doing so will keep data scientists abreast of the latest trends in the industry and update SEO-related companies’ machine learning stacks.
There are various journals and conferences such as IEEE International ConferenceAbout Machine Learning and Applications, to help you learn more about the latest machine learning trends.
It’s not directly related to SEO, but will help you understand the technological advancements that will disrupt your space next.
More resources:
Featured Image: Courtesy of Naveed Janvekar
