Today we are going to interview you Kumar Abhishek who works at Expedia Group, Inc as Machine Learning Engineer.
1. Please tell us a bit about yourself and your background?
I grew up in India and completed my master’s in Computer Science from the University of Florida, Gainesville. I always had a knack for learning new programming languages and using the latest technologies and frameworks to build and automate something.
I worked as a Software engineer at Amazon, a bay area startup, and at Microsoft for 6+ years. Then I decided to specialize in the field of machine learning and chose to be a machine learning engineer at Expedia Group, Inc.
2. What inspired you to switch to AI/ML space from being a software engineer?
One reason was to specialize and be an expert in one field but the bigger reason was taking up a new challenge. While working as a software engineer, I was taking up courses in machine learning from Coursera, a local university, and several other online courses and I absolutely loved how it blended math, programming, and research.
Having a software engineering background made me most suitable for the position of machine learning engineer and I switched teams internally within my company.
3. How is machine learning engineering different from traditional software engineering?
Machine learning engineers deal with everything that a typical software engineer does, however, there are two new entities or artifacts: the models and the data that is used to train the models. A lot of effort goes into monitoring the data quality, and usually, this data is huge(in terabytes) so big data processing frameworks like Apache Spark are used. The models are the end result of the training process, so monitoring and debugging model performance are where a lot of engineering time is spent.
4. Please tell us about your experiences in building machine learning systems or systems that leverage AI/ML.
I work on building and maintaining the machine learning pipeline for my team, which helps in detecting and preventing fraud for various brands under Expedia Group, Inc. This requires knowledge about big data technologies like Apache Spark as well as the inner workings of various machine learning models.
Most of the time, my work involves adding new capabilities to the machine learning pipeline or debugging any problems that may occur during the execution of the pipeline.
5. Lastly, what advice would you give to aspiring people trying to break into the machine learning space?
Immediate advice for people who are already working as software or data engineers is to try to identify an opportunity to use machine learning in whatever you are working on currently: see if you can become the go-to machine learning person in your team. If there is no such opportunity, then try switching to a team internally within your organization to enable this.
At the same time, keep getting your hands dirty by working on personal projects that leverage machine learning in some capacity, participating in online data science competitions on Kaggle or other similar platforms, and taking online or in-person courses. This would get your foundations strong and set you up for a great start.
Opinions expressed here are of Mr. Abhishek, not his employer