Machine learning forms an important segment of AWS. To foster more innovation and development in this area, AWS has established a Machine Learning Solutions Lab, which helps customers build machine learning solutions to address some of their pressing challenges. The ML Solutions Lab works backwards and delivers a roadmap of ML use cases and their implementation. In particular, the ML Solutions Lab offers solutions for personalisation and recommendations, computer vision, fraud prevention, and supply chain optimisation, among others.
Analytics India Magazine caught up with Priya Ponnapalli, senior manager at Machine Learning Solutions Lab.
ML Journey and AWS
“I have been working with Amazon ML Solutions Lab for three and a half years. This lab works with customers to help them adopt ML in their business. We brainstorm with them, identify their highest-value use cases, partner machine learning experts with these customers and help implement them. We work with customers all across the industries — healthcare, manufacturing, finance, media and entertainment, sports — on very many kinds of problems and really help their businesses move forward,” said Ponnapalli.
Ponnapalli told us that right from the start of her career, her focus has been implementing machine learning solutions to real-life situations and working at the intersection of different and diverse disciplines like cancer cure, art, pharma, finance, etc. “The opportunity to work with a breadth of customers working in different fields is what attracted me most about joining AWS,” she said.
A graduate in Electronics and Communication engineering, Ponnapalli first came to the US for her masters. She enrolled for MS in Electrical and Computer engineering at the University of Texas, Austin and went on to pursue her PhD at the same university. It was during this time that Ponnapalli worked at the Genomic Signal Processing Lab, which reaffirmed her interest and passion in the field of machine learning.
“I loved signal processing, specifically genomic signal processing. Applying signal processing and computational mechanisms to genomics that was coming from high throughput technologies intrigued me. In fact, I took that class in my first semester at the University of Texas. Here is where I found my love for applying computational mathematical techniques to data coming from different sources. This began my journey to data mining and machine learning. Throughout my PhD, I developed algorithms to work on large scale data for unsupervised learning models. I have always been interested in real-world impact. So, I decided to pursue a career which would allow me to apply ML to various applications,” said Ponnapalli.
Of the various projects that Ponnapalli is handling at AWS, she finds the ML solutions applied to sports most fascinating. Currently, her team is working with some of the largest sports organisations in the world, including the National Football League, Formula 1, Six Nations Rugby, and Swimming Australia, among others. Citing an example of AWS’ work in this area, Ponnapalli said, “One of the best examples is our work with NFL on player health and safety, where we combined NFL’s vast trove of data and their deep expertise in football with AWS’ ML and cloud computing expertise. We are working to transform sports, and we are doing this by analysis of player injury, game rules, rehabilitation and recovery. All this data will eventually be open to researchers, equipment manufacturers, trainers, coaches and medical professionals to serve as the framework of future innovation. I see this going beyond football to other sports and eventually be available in general.”
Present and Future of ML
“We are in the golden age of machine learning. We see businesses going from piloting ML projects to it having an impact on production. Many use cases of ML are in production that is helping companies stay competent and resilient. It’s no longer the question of whether you should have an ML strategy but how quickly can put their ML strategy to action,” believes Ponnapalli.
When asked about how companies should approach and build their data strategy, Ponnapalli listed four important points:
- Getting the data in order.
- Identifying the right ML use cases that deliver the most value for that business.
- Developing the culture of ML in their organisation — this could be through upskilling workforce and educating both technical and business leaders.
- Embracing the culture of ML, where iterating on things, embracing failures and repeating to see that transformation is at the crux.
Her advice to aspiring machine learning engineers — persevere and practice continuous upskilling. Her top resources for getting started are — Neural Networks and Deep Learning by Michael Nielsen, Dive into Deep Learning, and Deep Learning with Python by Francois Chollet.