Exclusive Q&A: Walgreens utilizes AI for vaccine scheduling, IT support
Walgreens is leveraging artificial intelligence (AI) in both internal- and external-facing areas of its enterprise operations.
Chain Store Age recently spoke with Yael Gomez, VP, Global IT, Integration and Intelligent Automation at Walgreens, about how the drugstore giant has been using the Digitate AIOps solution to streamline both vaccine scheduling for customers and managing tech support tickets from its IT department.
What kind of issues were you having with COVID-19 vaccine scheduling, and how did you resolve them with AI?
If you go back in time to when the COVID-19 vaccine was first made available, we were still having a very high number of COVID-19 cases in the country. One of the challenges for Walgreens was basically to make sure we could provide the right information to our customers as to what pharmacies would be open. Just like anybody else, Walgreens employees were getting sick with COVID-19.
[Read more: CVS, Walgreens pharmacies to begin administering COVID-19 vaccines]
Predicting or telling our customers when a store would be open and when the pharmacy would be available in the store suddenly became more difficult. The usage of the AIOps solution was really about making that information available in real-time. As we were learning the situation on the ground across all 9,000 stores, this information about stores and hours of operation could be updated in a timely fashion.
This was just so customers could know whether or not they could come to their local store and whether the vaccination service would be there. It was quite essential in the way we supported our store communities. We were able to divert customers to different stores. I think we shortened the time it takes to make that information available to customers threefold.
[Read more: Walgreens launches new IT model to save costs]
Are there any future plans or use cases for this technology that you can discuss?
Walgreens is pushing forward with a capability which is called workload management. If you think about how IT systems are run and managed, usually you find a number of coordinated processes happening and in the background, a.k.a batch, hundreds of thousands of these processes a month updating, aggregating, consolidating and moving data across multiple applications.
The successful and timely completion of these batches can be very critical to our operations and they are usually managed by specialized tool called an enterprise scheduler. With execution time varying depending on the month, day of the week, or volume of data to process, what the AIOps workload management capability is doing for us is giving us the ability to forecast, for instance, when a schedule will finish or what is the critical path to complete the work, enabling our teams to make appropriate decisions in support of the business.
For example, let's say we are processing insurance claims coming from patients whose volumes can vary enormously from one day to another. With workload management, what we are able to do now is predict how long it will take to run more batches and batch sequences, and therefore predict availability of certain information for the business, or when the system will be available for users to jump back in and do the rest of the work.
Another capability we’re exploring now is everything around performance management and capacity planning. This is an existing pain point in IT operations. It's hard to be good at performance management and capacity planning, you usually don't have enough metrics aggregated over a long enough period for you to understand nature and seasonality of the workload goes through the systems.
What we are exploring now is to basically inject all those performance metrics into the tool, let the tool learn from them, and pivot into basically understanding and forecasting what will be the demand in CPU, in memory, or any compute resource you want to measure.
Six months out, I can know if I have to provision something, have to recalibrate, reassess certain allocations of resources, and those sorts of things. With Walgreens’ scale of more than 9,000 stores nationwide, we have a significant amount of compute resources supporting business; applying the techniques of AI to our IT operations brings a huge benefit.