As the director of business intelligence and analytics at Dillard’s, Maiga oversaw an effort to use IBM Watson for forecasting product replenishment and allocation to the department store chain’s nearly 300 stores, as well as for deep customer analysis to support personalized marketing communications. Dillard’s now uses AI to make quick, accurate inventory decisions and to enhance the experience of its customers. With several recurring marketing campaigns, for example, Dillard’s significantly reduced its send volume, while increasing revenue.

What has your company achieved with Watson?
We conducted a proof of concept exercise in the fourth quarter of 2018 and signed a contract with IBM by the end of the year because the value of the Watson technology was demonstrated so quickly. We have models that could not complete on our old server that can now be run in a matter of minutes. We are also able to use deep learning modeling techniques, especially for image classification, that would have been impossible just a year ago. Not only are our customers receiving more relevant messages, but the number of inventory decisions being made each day would have required our merchants to be making a decision every 1.99 seconds.

What advice would you share with others who are considering using AI?
Prove the value of AI by starting with something small. Learn your business cases and find an advocate who really understands the current challenges. Then sell the capability, not the technology. 

What makes an AI project successful?
First, hiring the right people for your team. Look for people with not only the technical and data skills, but with a passion for solving problems and an ability to truly understand the business problem. Data is a lot of fun for data people, but not very meaningful to a business if it isn’t solving a direct problem, influencing a decision or producing something measurable. Second, picking a project that has very measurable metrics. Third, collaboration is key. The combined effort from our Data Science and Infrastructure teams has really melded the two teams together and made the journey look seamless to those benefiting from the results. 

What’s the best career advice you’ve received?
Whenever you think or hear “we can’t because...,” rephrase it and use “we can if....” Once you get on paper what would be required to achieve what is being asked, no matter how unattainable it seems, you can then budget and prioritize and make informed decisions on whether to move forward. Most things that seem impossible are quite doable once you’ve thought through all the options and requirements. Also: If there are 10 things that are stressing you out and you have control or influence over eight of them, focus on those eight and let the other two go.

She puts Watson Machine Learning Accelerator to work

Dive deeper into your journey to AI and learn more about Watson Machine Learning Accelerator.