In 1955, NBC President Sylvester Weaver delivered a speech to the Institute of Practitioners of Advertising in London. The speech, subsequently reproduced in the Daily Iowan, addressed what Weaver diagnosed as the beginning of an “information explosion.”
“To each man there is flooding more information than he can presently handle,” said Weaver nearly 70 years ago, “but he is learning how to handle it and, as he learns, it will do him good.”
The ensuing decades have certainly proven Weaver prescient — if not downright premature in scope of his assessment. Data is more ubiquitous than it’s ever been. In just the first five years of the 2020s, the International Data Corporation predicts that digital device users will produce more than twice the amount of digital information created since the advent of digital storage in the 1950s.
What’s more, anyone with a passing knowledge of Facebook’s business model might tell you that data is valuable. But it’s the ability to sift away noise and isolate that which is salient in this profusion of data that is truly prized. Enter the data scientist — sifting chaff from those valuable kernels of information. If this unprecedented proliferation of data is to do us good, as Weaver predicted, it will be through the hard work of such tech professionals.
“Data is so ingrained into our life at this point I don’t think there’s a single industry that’s not touched by data,” Michelle McSweeney, data science domain manager for Codecademy, recently told ZDNet in an article on growing interest in the data science role. “There’s not a single company that’s not touched by data.”
More than 2.3 million users have enrolled in CodeAcademy’s data science career path in the last two years, with jobseekers hoping to fortify their skills and enter the burgeoning data science job market.
Built In NYC sat down with four scientists working on the forefront of data science to learn more about their roles, their teams and the ample rewards of working in the field.
From its inception servicing F1 racing teams to its current iteration as the artificial intelligence arm of management consulting firm McKinsey, QuantumBlack has served clients and geographies around the world. The company employs “hybrid intelligence,” using AI in conjunction with strategic thinking and domain expertise, to analyze data and deliver insights to its customers.
What is the coolest project you’ve worked on or are working on as a data scientist at QuantumBlack?
One of my favorite projects was using advanced analytics to help an online listing business develop a growth strategy and customer personalization. We built a customer 360-view using multiple data sources including clickstream and listing information, as well as post-lead responses. We were using analytics to bridge the gap between the business and technical side of the organization, untapping actionable insights and improving the customer experience at the client’s website. The mix of solving both technical problems — like the identity resolution of web visitors — and business problems — such as addressing how to better market to sub-segments of customers — is truly the most rewarding part of the project.
More generally, what excites you about the work you’re doing with QuantumBlack?
Personally, it excites me to think about the problems we are working to solve, which is the beauty of being a technical consultant and acting as a thought partner to our clients. Most of the time, the problem-solving space involves a mix of technology, data and analytics and people elements including capabilities and team set up. It excites me to work with the clients side by side, learn how they think about their job and business and identify the most suitable analytics solution, which eventually our clients will own.
It excites me to work with the clients side by side.”
What do you love most about the culture of QuantumBlack?
I like that we celebrate each other’s strengths at McKinsey and QuantumBlack. We all have something new to develop personally, but it’s equally important that we play to our strengths and contribute to the teams.
Identifying my strengths and learning how to best use them has always been the key element in the feedback sessions. It enables me to bring my authentic self to work.
A data-science driven marketing and market research company, Known pairs data scientists with award-winning creatives, expert research teams and strategists to develop advanced, end-to-end solutions for its clients.
What is the coolest project you’ve worked on or are working on as a data scientist at Known?
At Known, our proprietary suite of technical tools is called Skeptic. It helps our data scientists and media buyers drive advertising-based user insights, automate manual processes and make strategic decisions as an agency.
One of the coolest tools helps us find the best keywords for search campaigns. Whenever someone searches for a term, we have an opportunity to bid on an ad impression. Deciding which terms to bid on is a complicated problem. There is a huge “white space” of possible terms, and we can’t possibly think of all of them – let alone test them all.
Using natural language processing tools, we built a means to visualize and map out possible keywords while incorporating past performance and scalability. We identify what is working, what isn’t and where we are missing keywords that have a high likelihood of success. We have paired tools like this with models that predict how keywords would perform based on their semantic content to quickly optimize and drive campaigns forward. This, combined with the extremely talented search analysts on our team, results in relentlessly optimized campaigns.
More generally, what excites you about the work you’re doing at Known?
Advertising is a big data problem, but many of the challenges we are working on don’t have out-of-the-box solutions. Our data scientists need to think outside the box to identify how to use the data we have to build the solutions we need. We train them to be media experts as well as highly capable technical and quantitative practitioners. This mentality drives them to develop tools and processes that deliver maximal impact.
We are always asking ourselves questions and trying to find the right answers.”
When do you love most about the culture of Known?
At Known, everything starts with science. I love the inquisitive and collaborative nature that we foster on our team. One of our core values is, “There’s always a better way.” We are always asking ourselves questions and trying to find the right answers.
Because we have both technical and non-technical people on our media-buying team, we all bring to the table different areas of expertise. This means we can take advantage of each respective person’s strength to build the best tools possible. We look to bring people in who offer a different skill set, truly enjoy helping each other and want to share their knowledge.
For more than 160 years, Northwestern Mutual has been servicing clients’ insurance and investment needs. The company serves roughly 4.5 million customers, holds $265.0 billion in assets, takes in $28.1 billion in yearly revenues and insures $1.8 trillion worth of life insurance protection worldwide.
What is the coolest project you’ve worked on or are working on as a data scientist at Northwestern Mutual?
I work on the mortality model team, which is part of the larger underwriting team at Northwestern Mutual. Our objective is to model our client’s mortality risk to better understand the drivers, causes and relationships that are associated with mortality risk. This project has brought about unique theoretical and technological challenges — the kind of challenges that require in-depth collaboration across the organization, as well as the need to play around with and experiment with technology.
More generally, what excites you about the work you’re doing at Northwestern Mutual?
We’re given the creative freedom to experiment. The excitement comes from having smart people around you who you can bounce ideas off and collaborate with. It comes from having the right technology at your disposal. Work becomes a rewarding series of challenges that are increasingly complex and fun. The best part is that with each new challenge you grow as a person, and that sense of growth and accomplishment can’t be beat.
With each new challenge you grow as a person, and that sense of growth and accomplishment can’t be beat.”
When do you love most about the culture of Northwestern Mutual?
One of the greatest parts of the culture here at Northwestern Mutual is the shared enthusiasm for the work we do, and our team’s regular collaboration. We really are invested in the work, and when you combine that with smart colleagues and a general passion for coming up with new ideas, you get a culture of brainstorming and partnership that I absolutely love.
When we run into a particularly tough problem, we bounce the problem around. We research it and play around with it. Inevitably someone suggests a dedicated brainstorming session. We block off a few hours on our calendar, and we whiteboard the problem. It reminds me of working on long problems in college and having peers that you can joke around with while getting through the tough work together.
Prescriptive Data employs operational and information technology to provide greater energy efficiency, cost savings and enhanced thermal comfort to customers in the real estate space. The company’s flagship product Nantum OS is a building management platform that uses data-driven insights from IoT, machine learning and artificial intelligence.
What is the coolest project you’ve worked on or are working on as a data scientist at Prescriptive Data?
In the past five years at Prescriptive Data, I have initiated and been involved in a broad range of energy and non-energy centric projects that we, in the PD research team, used novel AI architectures to formulate and investigate. These ranged from formulating a fully-automated building startup to collecting real-time occupancy data like human detection and counting using video surveillance cameras.
I think the current project that we initiated, greenhouse gas or GHG-based Distributed Energy Resource Management, is one of the coolest projects I have been involved in so far. The challenges we have encountered in formulating the AI architecture of this intricately interconnected complex system, comprising numerous subsystems and using physics-informed machine learning to model the behavior of different systems, make me and my team very excited. But, besides the complexity of this system of systems, the potential impact of this tool on GHG emission reduction, and eventually the decarbonization of buildings, make us thrilled to have the chance to work on this project.
More generally, what excites you about the work you’re doing at Prescriptive Data and the technology you’re using?
I have been involved in performing fundamental research around the application of machine learning, mathematical optimization, physics-informed modeling and artificial intelligence over the past decade. I’ve been chairing AI, optimization and mathematical modeling sessions at the American Society of Mechanical Engineers conferences since 2015. Working in PD Core Research is exciting as it gives us a unique opportunity of working in a fast-paced, production-driven environment with the objective of developing cutting-edge, robust and reliable AI- and ML-based industry-level applications.
It is very interesting seeing how AI technologies not only offer the opportunity to significantly reduce carbon emissions (through robust and reliable automation control) in smart buildings but also provide a more pleasant environment for tenants.
Our product, Nantum OS, is incredible because of the caliber of people who work on it behind the scenes.’’
When do you love most about the culture of your team at Prescriptive Data?
When we started the core research function at Prescriptive Data, we intentionally defined its culture like those we had experienced and explored in the most successful university research teams. I tried to manage the function based on what I experienced at Columbia University.
With that in mind, both our core research team and other departments in Prescriptive Data are staffed with intelligent and hardworking people. When approached with an opportunity to work closely together, in circumstances like AI-intensive hackathons, my team delegates, meets deadlines and often wins these public challenges. Our product, Nantum OS, is incredible because of the caliber of people who work on it behind the scenes.
