An interview with Andrew Ferlitsch, Developer Program Engineer at Google
Our interviewee today is Andrew Ferlitsch. Andrew is a Developer Program Engineer at Google where he develops curriculum and teaches artificial intelligence, data science, machine learning, and computer vision.
Formerly, he was a Principal Research Scientist and a Project Manager with over 20 years of experience in research fields and 30 years of industry experience, with 115 US issued patents. He holds a Masters in Computer Science, with a major in Artificial Intelligence. He has worked a wide range of technology fields for commercial driven research. His primary expertise in the last 5+ years has been on machine learning, neural networks, geospatial technologies, data science/analytics, open/big data, and autonomous vehicles.
Andrew is also the author of the Idiomatic Programmer handbook series on Keras and Computer Vision which is very focused on developers in general. I highly recommend checking it out.
I would like to wholeheartedly thank Andrew for taking the time to do this interview. I hope this interview serves a purpose towards the betterment of data science and machine learning communities in general :)
An interview with Andrew Ferlitsch, Developer Program Engineer at Google
Sayak: Hi Andrew! Thank you for doing this interview. It’s a pleasure to have you here today.
Andrew: Sayak, thank you for inviting me.
Sayak: Maybe you could start by introducing yourself — what is your current job and what are your responsibilities over there?
Andrew: I work for Google in Developer Relations for Machine Learning. My primary responsibility is reaching out to software development communities, universities and high school STEM programs, as an AI/ML/DS educator.
Sayak: That’s awesome! Would you like to share how did you become interested in data science and machine learning?
Andrew: I’ve always been a data scientist in one way or another. I got a graduate degree in AI in 1987, but back then was the AI Winter. I ended up as a research scientist working in Japanese IT, specializing in imaging/computer vision and data. So it was a natural hop to move into today’s definition of data scientists and machine learning.
Sayak: I am glad that we share a similar subject of interest in computer vision. When you were starting what kind of challenges did you face? How did you overcome them?
Andrew: I think my biggest challenge was back in 2016–2017 having to “relearn” statistics. I hadn’t had a need in my career for advanced math. I pushed myself to take online courses, read books, etc until it was all refreshed in my head again. Today, I teach statistics along with other ML topics at universities as an adjunct professor.
Sayak: This is motivating, Andrew :) In fact, I have read your short handbook on Statistics (offered via the Idiomatic Programmer Handbook series) and it has helped me a lot in rethinking many of the concepts. Thanks to you for that! What were some of the capstone projects you did during your formative years?
Andrew: I wouldn’t say I have a capstone project, in that I was fortunate with my background of working on autonomous vehicles before ML, to be able to find freelance jobs and learn on the job. Some early ones involved document understanding using NLP with patient and insurance records in healthcare. Another was using computer vision to detect diseases and infestation in crops in real-time with an autonomous land drone, which can then direct an aerial drone to spray the area infestation in real-time.
Sayak: Those are some really fantastic applications you mentioned there. These fields data science and machine learning are rapidly evolving. How do you manage to keep track of the latest relevant happenings?
Andrew: Read, Code, and Teach. I read lots of research papers and then try to validate their results.
Sayak: I too reciprocate with you on this philosophy of reading, coding, and teaching. I think teaching always helps you learning new things and pushing your limits. Being a practitioner, one thing that I often find myself struggling with is learning a new concept. Would you like to share how do you approach that process?
Andrew: For new concepts, I start with the foundational work in the research papers. Then I modify what was in the paper and see what happens. I call those my CrazyNets. Finally, I find teaching helps keeps new concepts fresh in my mind.
Sayak: That was a great input, thanks, Andrew. The Idiomatic Programmer handbook series is such a gem of a resource. Would you like to share what motivated you to start working on it?
Andrew: When I first started teaching deep learning, most of my students came from a statistics background, but did not know how to code. As time evolved, that shifted to students who were software engineers but didn’t know statistics.
To overcome teaching both groups, I developed the idiomatic teaching methodology for the purpose of AI/ML/DS education being accessible and inclusive to all.
Sayak: I am sure this had helped the community to a great extent. Is there any way the Machine Learning community can contribute to this project?
Andrew: Yes. I think the best way to start is by reaching out to me. I can discuss with the person how they like to contribute and give them guidance.
Sayak: Duly noted, Andrew. Any advice for the beginners?
Andrew: AI/ML/DS has moved at a very fast pace. I don’t think self-learning will work anymore, like back in 2016–17. I would recommend that one starts with an accredited course. If you are already a strong Python programmer, I recommend a 6 to 9 months’ Nanodegree.
Sayak: I am sure many will find this advice to be extremely helpful. Well, thank you so much, Andrew, for doing this interview and for sharing your valuable insights. I hope they will be immensely helpful for the community.
Andrew: Again, Sayak thank you for inviting me and it is a pleasure to provide your community with whatever insights I can offer.
Summary
Andrew did not give it a second thought to relearn the statistical concepts in order to get himself a strong grip on the subject. He pushed himself for that and now he is an adjunct professor at a number of universities teaching statistics along with other machine learning topics. Isn’t that amazing? Isn’t that something that we all should learn — to keep on learning things?
I hope you enjoyed reading this interview. Watch out this space for the next one and I hope to see you soon. This is where you can find all the interviews done so far.
If you want to know more about me, check out my website.
