Consumers most likely don’t understand how giant a task data science performs in retail. The self-discipline gives details about shopper habits to assist predict demand for merchandise. It’s additionally used to set costs, decide the variety of gadgets to be manufactured, and determine extra environment friendly methods to move items.
These are simply a number of the insights that knowledge scientist Vivek Anand extracts to tell choice makers on the Gap, a clothes firmheadquartered in San Francisco. As director of data science, Anand—who is based in Austin, Texas—manages a team that includes statisticians and operations research professionals. The team collects, analyzes, and interprets the data, then suggests ways to improve the company’s operations.
“Data science is trying to effectively solve problems that were previously unsolvable,” Anand says. “The technology is used to group similar transactions that look different on the surface. But underneath they are similar.”
Anand is an IEEE senior member who has spent his career using data science, artificial intelligence, and mathematical and statistical modeling to assist companies resolve issues and make smarter selections.
Final yr AIM Research honored Anand’s efforts to rework the retail trade with its AI100 award, which acknowledges the 100 most influential AI leaders in the US.
An information scientist at coronary heart
Rising up in Gopalganj, India, he set his sights on turning into a doctor. In 2006 he enrolled within the Indian Institute of Science Education and Research (IISER) in Pune with each intention of incomes a medical diploma. Throughout his first semester, nevertheless, he loved the introductory arithmetic lessons way more than his biology programs. A challenge to design a statistics program to find out the easiest way to vaccinate individuals (pre-COVID-19) helped him understand math was a greater match.
“That was my first introduction to optimization methods,” he says, including that he discovered he actually appreciated figuring out whether or not a system was working as effectively as doable.
The vaccine challenge additionally bought him thinking about studying extra about industrial engineering and operations analysis, which makes use of mathematical modeling and analytical methods to assist advanced programs run easily.
He graduated in 2011 from IISER’s five-year twin science diploma program with bachelor’s and grasp’s levels, with a focus in arithmetic. He then earned a grasp’s diploma in operations analysis in 2012 from Columbia.
One of many programs at Columbia that intrigued him most, he says, was enhancing the method of figuring out an individual’s danger tolerance when making funding selections. That coaching and an internship at an funding agency helped him land his first job at Markit, now part of S&P Global, a credit-rating company in New York Metropolis. He created AI and mathematical fashions for monetary transactions corresponding to pricing money and credit score devices, together with credit score default swaps. A CDS is a monetary instrument that lets buyers swap or offset their credit score danger with these from one other investor.
Anand, who started as an analyst in 2013, was promoted to assistant vp in 2015.
Later that yr, he was recruited by Citigroup, an funding financial institution and monetary companies firm in New York Metropolis. As an assistant vp, he developed knowledge science and machine learning fashions to cost bonds extra precisely. He additionally led a workforce of quantitative analysts accountable for modeling, pricing, and figuring out the valuation of credit score derivatives corresponding to CDSs in rising markets.
He left Citi in 2018 to hitch Zilliant, a worth and income optimization consultancy agency in Austin. As a senior knowledge scientist and later as lead knowledge scientist and director of science, he led a workforce that constructed and serviced customized worth optimization fashions for patrons within the automotive, electronics, retail, and meals and beverage industries.
“We used to estimate elasticities, which is a key element for pricing merchandise,” he says. Worth elasticity exhibits how a lot demand for a product would change when its price modifications. “The present algorithms weren’t environment friendly. In a lot of situations, it used to take days to compute elasticities, and we have been in a position to carry down that course of to some hours.”
He was director of science at Zilliant when he left to hitch the Hole, the place he oversees three knowledge science subteams: worth optimization, stock administration, and achievement optimization.
“Within the style trade a overwhelming majority of product assortments are repeatedly refreshed,” he says, “so the target is to promote them as profitably and as shortly as doable.” Clothes tends to be season-specific, and shops make area on their cabinets for brand spanking new gadgets to keep away from extra stock and markdowns.
“It’s a stability between being productive and worthwhile,” Anand says. “Pricing is mainly a three-prong method. You need to maintain onto stock to promote it extra profitably, clear the cabinets if there may be extreme unproductive stock, and purchase new prospects by means of strategic promotions.”
Managing stock will be difficult as a result of the vast majority of style merchandise bought in the US is made in Asia. Anand says it means lengthy lead instances for supply to the Hole’s distribution facilities to make sure gadgets can be found in time for the suitable season. Sudden delivery delays occur for a lot of causes.
The important thing to managing stock is to not be overstocked or understocked, Anand says. Information science not solely can assist estimate the typical anticipated supply instances from totally different international locations and consider delivery delays but in addition can inform the optimum portions purchased. Given the lengthy lead instances, correcting an underbuy error is tough, he says, whereas overbuys lead to unsold stock.
Till not too long ago, he says, specialists estimated transit time based mostly on common supply instances, they usually made educated guesses about how a lot stock for a sure merchandise could be wanted. Generally, there is no such thing as a definitive proper or fallacious reply, he says.
“Primarily based on my observations in my present function, in addition to my earlier expertise at Zilliant the place I collaborated with a variety of organizations—together with Fortune 500 corporations throughout varied industries—knowledge science fashions often outperform material specialists,” he says.
Constructing an expert community
Anand joined IEEE final yr on the urging of his spouse, laptop engineer Richa Deo, a member.
As a result of knowledge science is a comparatively new area, he says, it has been tough to discover a skilled group of like-minded individuals. Deo inspired him to contact IEEE members on her LinkedIn account.
After many productive conversations with a number of members, he says, he felt that IEEE is the place he belongs.
“IEEE has helped me construct that skilled community that I used to be on the lookout for,” he says.
