With a dozen years at Amazon under his belt, Deepak Bhatia made a big career change last year when he joined Hershey as the candymaker’s first-ever chief technology officer.
“The goal is not to simply take things that were at Amazon and bring them over,” says Bhatia, whose last role at the e-commerce giant was vice president of supply-chain optimization technologies.
He asserts that not all of the technology applications that worked at Amazon, which has expertise in cloud computing, robotics, and generative artificial intelligence, can also be applied to Hershey—a consumer packaged goods company that produces, manufactures, and sells the namesake candy bar, Twizzlers, Kit Kat, and other snacks.
But there are lessons learned at Amazon that he does plan to infuse at Hershey, particularly around automating tasks within the foodmaker’s manufacturing supply chain and inventory management. Some other areas of focus for Bhatia as CTO include investing in data science, forecasting, cybersecurity, and more precise spending on marketing and in-store promotions.
As Hershey looks to rely more on data for those decisions, which at times may be automated, humans will remain firmly in the loop. “We have to balance it with a measured level of skepticism in the machine output and not just blindly trust it,” says Bhatia.
He’s particularly excited about how gen AI can help generate marketing assets and develop new snacking innovations, while also bolstering productivity with AI-enabled agents, copilots, and assistants to take on more mundane tasks.
To get the workforce up to speed, Bhatia says Hershey offers tech courses online, but a bigger transformation is underway within the technology group, where there is less focus on a project-by-project approach and more emphasis on empowering tech leaders to collaborate and create solutions for the business.
While Bhatia is truly excited about gen AI, his approach is not to throw the technology at everything. He wants to start with simple solutions and add complexity when it is justified: “That’s how I’ve seen a lot of these new technologies generate a lot of value over time.”
This interview has been edited and condensed for clarity.
You are the first-ever CTO at Hershey. What led them to create the role, and what were the priorities you laid out?
We’ve had CIOs before, but the first CTO is a strong message and gives the status to the role that it deserves in terms of what we want technology to do. First and foremost, with our people and talent, and what should be in house [versus] what should be external. Our preference is going to be investing in talent and capabilities that will future-proof us. Second, which I think is underestimated, is the value of data. The focus on data—making good quality data accessible and modernizing the tech—this is where our investments are going to be.
How are you remaking the company and the tech teams to help make a difference across the business?
Core to this is obviously having the right talent. But how we organize is equally important. We are working very actively to reorganize, and that’s obviously going to be a journey. That can be anything from how we do end-to-end planning, including sales and operations, to supply chain. Other examples can be, how do we better spend and optimize our trade promotions and media spending and get the most bang for our buck? And the third is a never-ending desire to better understand the customer and better understand the categories and competition.
What is the status of Hershey’s generative AI journey?
In the near term, there is huge potential that gen AI has in terms of agents, assistants, and pilots that really augment what people can do, automate it, and make it faster. We are investing a lot in gen AI in the application of how we do promotions, marketing, and content generation. It is the beginning of the journey, and the effects of this will manifest in the years to come. We are making huge investments not only in terms of technology, but also how we organize ourselves and how we invest in talent.
Do you have generative AI in production today or are you still in test and pilot mode?
We are starting to test and pilot a lot of gen AI applications at the same time. If you have gen AI being used as a pilot in getting ideation and marketing content generation, then I would consider this to be more than a pilot. But it’s hard to say how much of a role that particular work stream you can attribute to gen AI versus what the actual experts are working on. Keep in mind, the sweet spot for a lot of these applications and technologies right now, especially on the generative side, is how they can augment or scale the ideation process.
What do you think about the cost of generative AI and balancing those costs with the outcomes you want to achieve?
There have been huge investments that have been made by a lot of big companies in gen AI. The monetization hasn’t yet been completely figured out. What you will see is that there will be a few iterations of a pricing model and different solution providers will experiment and figure out a really effective way to monetize. It has to happen with the amount of investments that have been made on this entire ecosystem. But I think it’s very early to know how that will happen, and likely a one-size-fits-all strategy won’t end up manifesting.
What is your favorite Hershey snack?
I love our dark chocolate Kit Kat. But there’s a Hershey golden almond bar that’s only available in Hershey’s stores. If you get a chance to go to one of our physical stores, get one of those dark chocolate almond bars. I buy boatloads of them and bring them back home.