[ad_1]
Two hockey players face off over the puck. Which one is likely to prevail? Fans can now get a real-time tip from a new National Hockey League stat, Face-Off Probability, powered by Amazon Web Services (AWS). The in-game hockey probability is just one of many ways machine learning is showing up in daily life, from sports to healthcare to finance.
The possibilities for AWS customers are vast, and many of those ideas are cutting-edge ones that have never been tried before. As director of applied science at the Amazon ML Solutions Lab, Priya Ponnapalli leads a global team of scientists, engineers, and product managers who help AWS customers identify and implement their most important machine learning opportunities.
The demand for her team’s expertise is large and growing, according to Ponnapalli, who joined the lab as a principal scientist nearly four years ago. Other AWS sports partnerships include helping to create the National Football League’s real-time Next Gen Stats and new design specifications for F1 race cars.
Discovering machine learning innovations
Sports leagues make up one segment within a broad portfolio of ML Solutions Lab customers that includes automotive, manufacturing, healthcare and life sciences, and financial sector customers.
“We’ve got nearly 100 engagements with customers globally across all industries, and no two engagements are alike,” Ponnapalli says. Her current focus: building a scalable organization needed to support increasing AWS customer demand.
“One of the most important things I do is invest in growing my talent and growing my leaders,” she says. “I’m always encouraging my team to think big, look around corners for what’s possible, and anticipate how best we can serve our customers.”
Last year, Business Insider named Ponnapalli to its list of 100 people transforming business, recognizing her work in leading businesses into the machine learning landscape.
“I got included along with many heroes that I admire and I just felt very honored,” Ponnapalli says. “Of course, it’s reflective of all the good work that my team does.”
An ML Solutions Lab relationship often begins with a discovery workshop, Ponnapalli explains, where the customer shares what their biggest challenges and opportunities are, as well as what kind of data assets they have. That workshop then informs use cases for the AWS team to envision and build.
“Sometimes customers come to us with an open-ended charter. Others come to us with a very specific problem, like, ‘We want to detect acoustic anomalies to monitor equipment performance and anticipate failures on our manufacturing floor,’” she says. “It varies, but our process is rooted in Amazon’s working backwards philosophy.”
From math to machine learning
Growing up in Hyderabad, India, Ponnapalli always loved math. Because she went to an all-girls school, she says, there was no stereotyping about what one gender could or could not do. Her parents instilled confidence in her as well.
“I’ve had a lot of strong role models throughout my life. They’ve all been very inspirational and always made me feel like I could do anything that I wanted to,” she says. “I want to be able to create that same environment for others through my leadership of my team.”
After earning her bachelor’s degree in electronics and communications engineering at Osmania University in Hyderabad, Ponnapalli went to the University of Texas at Austin, where she got her master’s and doctorate degrees in electrical and computer engineering.
Coming out of undergrad, she was very interested in digital signal processing. A class in graduate school on signal processing for genomics turned out to be pivotal for her career; the professor of the class, Orly Alter, invited Ponnapalli to join her lab and became her PhD thesis advisor.
Alter’s course was also Ponnapalli’s introduction to machine learning. Building on that foundation and her thesis, she and Alter co-founded Eigengene, which uses artificial intelligence to analyze cancer genome data and create personalized diagnostics and prognostics. They developed algorithms designed to find patterns within diverse, high-dimensional datasets known as tensors.
As she conducted research related to tensor decompositions, Ponnapalli stumbled across a lot of finance forums where the same questions were being asked. “I realized that these algorithms are data-agnostic. They are industry-agnostic, with broad application within multiple areas,” she says.
She dropped off her resume with Bloomberg at a UT Austin career fair late one week, interviewed over the weekend, and got a job offer that Monday. At Bloomberg, she developed a social media analytics tool called Bloomberg Social Velocity that alerted clients to spikes in social activity and market sentiment about companies. Lead data science roles at JP Morgan Chase and Genentech followed.
“I’ve always enjoyed working at the intersection of multiple disciplines,” Ponnapalli says. “I truly believe that’s where the magic happens, where there’s a lot of cross-pollination of ideas.”
The opportunity to work across different industry sectors while maintaining a focus on machine learning led Ponnapalli to apply for a position with the ML Solutions Lab. In addition to her role at Amazon, Ponnapalli teaches machine learning to business leaders as faculty at Rutgers University.
“I’m most excited about solving problems with real-world impact,” she says. “Machine learning has the potential to help us solve some of the most complex and challenging problems of our time, like cancer and climate change.”
She notes that with so many opportunities in machine learning, from scientists to product managers to engineers, anyone with an interest should not be afraid to pursue a job in the field. Given the availability of online resources, she adds, it’s never been easier to learn. Plus, the field needs a diverse pool of talent.
“Creating inclusive workspaces is a cornerstone of my leadership style. I’ve used my position and whatever privilege and success that has come to me to help others and hire a diverse team,” she says. “This is a field that impacts all of us — the products that we build need to work for everybody.”
[ad_2]
Source link
