Sticky problems keep our data scientists engaged
By Sarmila Basu, Jul 06, 2017
Sarmila Basu (center) works with her Microsoft IT Data and Decision Sciences team.
When you bring together a wildly diverse group of geniuses, the hard part isn’t finding work for them to do; it’s finding something that’s hard for them to solve, something so challenging that they get a little bit mad and a lot fired up.
If not, they’ll get bored and they might wander off.
That’s why it has taken me seven years to build my team: an eclectic mix of statisticians, economists, mathematicians, electrical engineers, biophysicists, and telecommunications specialists who are helping shape the way Microsoft uses data.
When I started my team in 2010, the Microsoft IT Data and Decision Sciences Group, it was just a one-woman band. Each year I added two or three people, always making sure I had an appropriately challenging set of work for them to take on (we’re now up to 26 people). So far, no one has walked away bored, and together we’re charting a new path in data analytics.
One challenge of building this team has been finding the right people because, as it so happens, the kind of data science talents that we’re looking for are very hard to find. It’s almost like you’re looking for data artists because you want them to have scientific training, you want them to be able to talk to business leaders, you want them to be good storytellers, and you want them to be able to click with all the other unusual personalities on the team. Put simply, we require a multi-facetted skillset that’s often hard to find in just one person.
Recruiting people who look at life in massively different ways has challenged me in ways I never would have thought possible. Some people on my team are very quiet. Some are loudly vocal. Some like to burn the midnight oil, and some like to start working at the crack of dawn. We draw from people who grew up in different cities and cultures around the globe. We have people who are three decades into their careers and others who are just out of college. Some who have been at Microsoft for their entire careers and others who just got here. With such diversity, everyone has had to be willing to meet each other half way, creating a great, collegial atmosphere where we’re always learning.
The Medici Effect
Clearly one of the core principles in building out my team has been diversity, but when I say diversity, I mean diversity of discipline, not just demographics. What we are looking for is diversity of thought as well as experience.
I’m a big fan of The Medici Effect, a book that talks about how, at the beginning of the European Renaissance in the 14th Century, the Medici family of Florence, Italy, invested heavily in people from a wide variety of disciplines because they believed the best outcomes come from the confluence of multiple skillsets and multiple ways of thinking. They sponsored Leonardo da Vinci and many other creative types like him for those reasons. And today, this way of thinking is having its own renaissance. That’s why you see things like Harvard medical doctors fighting drug-resistant mutating AIDS cells using the same techniques computer scientists use to fight spam.
And that diversity isn’t just something we look for on hiring day–we use it every day because it’s exactly this kind of experiential diversity that leads to breakthrough ideas and innovative solutions.
When we helped the Tacoma School District in Tacoma, Washington, predict which kids were likely to drop out of school, we did it by pairing our architect with one of our mathematicians. They worked on a predictive model that enabled the district to reach vulnerable kids before they dropped out, turning one of the state’s highest dropout rates (42 percent) into one of its best (down to 28 percent in just one year).
When we helped ThyssenKrupp Elevator predict which of the 1.1 million elevators it maintains was going to break next, we did it with help from our PhD in electrical engineering and our PhD who has worked in manufacturing. Thanks to the big data-fueled model they built, some Internet of Things sensors, and tools like Power BI and Azure Machine Learning, the company could cut its maintenance bill. The costs are going down over time too, because the model we built for them learns over time.
When we helped the Skype team improve the quality of Skype calls, we did it by pairing our engineer who has a background in network engineering with our statistician who specializes in broken processes. Together they established a standard of what constitutes good call quality. Then we used decision tree models to isolate the issues that lead to poor user experiences for people using Skype. The project team could then identify combinations of factors that were causing problems in call quality and isolate scenarios where calls were breaking down. For example, they could determine which listening devices, locations, and networks were being used for further troubleshooting.
What does it all mean?
Tying all this together, what we’re doing is we’re using artificial intelligence, the cloud, and data analytics to create new services and products that didn’t exist before. Every day we’re learning and refining what we do with one clear goal in mind – we want to replicate our work in ways that allow us to help more than one customer at a time. And when we do that, we also want to find ways for Microsoft to monetize our efforts.
And we’re already putting this into action. We are taking the model we built for the Tacoma School District to nearby Bellevue College and we’re refining it as we go. And next time we’ll fine tune it a little more, and soon—if we do things right—we’ll be helping hundreds if not thousands of schools transform the way they keep their students from dropping out. When we start to scale our impact beyond one-off projects, we then begin reaching our true goal, which is to help our customers, and Microsoft itself, digitally transform.
Stay tuned over the coming months as I use this space to share more stories on how we are using data analytics to change the way our customers work, customers inside Microsoft and externally. Next time I’ll go deep into a specific example of data analytics work we did for one of our customers. Meanwhile, read my first post on how I started in this role and why Microsoft IT is home to my team.
This blog post was originally published in Microsoft’s IT Showcase.
About the author
Dr. Sarmila Basu is Chief Data Scientist at Microsoft IT, leads the MSIT Data Sciences (DDSG) group and oversees both internal and external global customer engagements. Dr. Basu is instrumental in Microsoft’s forward-looking development efforts to integrate the company’s Data Science expertise with the Sales, Services and Product Group. In this role, she has been heavily influential in the direction and execution of the company’s commitment to leveraging Advanced Analytics solutions using Microsoft technology. Using this Data Science strategy maximizes management’s ability to make data driven decisions and solve challenging business problems.
Dr. Basu is a Ph.D. in Economics, and has 20 years of senior management experience working with Fortune 500 companies in Telecom, Financial Services and IT. She is passionate about Data Science and the business value it delivers. Sarmila firmly believes there is an art to Data Science. It requires more than Advanced Analytics depth and technical expertise to make a material difference in how we help our company and customers.
Dr. Basu is a noted and sought after conference speaker. Work from Dr. Basu’s team is featured on Microsoft’s IT Showcase. You can find Dr. Basu’s writings on Linkedin. When not at work, Sarmila is committed to her philanthropic pursuits. These interests include Economic Empowerment and as a board member for the Arts.
Apply for the 2017 ANNY
The ANNY award recognizes organizations’ applications of advanced analytics to drive measurable business results at the project or organizational level. To apply for the ANNY, complete the application and email it to firstname.lastname@example.org.
Accelerate your organization’s journey to analytics maturity
Get the data sheet to learn how the Research & Advisory Network advances analytics capabilities and improves performance.