Allstate Chief Information Officer Zulfi Jeevanjee says his initial interest in technology is fortuitously due to a hijacking incident in Africa.
Born and raised in Kenya, Jeevanjee never used a computer as a child, but developed early aptitude in subjects like chemistry, physics, and mathematics. He had hoped to enroll in a British university, the path his parents took before him. But the armed truck carrying his exams was hijacked and never made it to Europe. The universities said he’d have to wait several months to apply.
Jeevanjee shifted his focus to America, where he earned a scholarship from Washington University in St. Louis. After taking an introductory computer science course, he switched his major to computer engineering. “I just love the fact that I could listen to someone’s problem and then just convert that into a solution,” says Jeevanjee.
Throughout his career of more than three decades, Jeevanjee held leadership roles at Wells Fargo as it digitized consumer banking and integrated the acquisitions of Wachovia and A.G. Edwards. Jeevanjee is a boomerang employee at Allstate, initially joining in 2018 to lead the architecture, innovation, and systems engineering teams, before leaping to CVS Health as chief technology officer during the pandemic, and then rejoining Allstate in 2022 as CIO.
“Allstate, like most insurers, was saying ‘Look, we got to change the way we distribute our products,” says Jeevanjee.
The industry was moving away from selling home, auto, and other insurance policies over the phone via agents—a bulk of Allstate’s business—to more directly selling insurance online. Transforming how insurance was sold also required a restructuring of the technology team, which relied on project managers and business analysts to prioritize the delivery of tech solutions on time and within budget.
Those project manager roles were eliminated and workers were retrained to focus almost exclusively on engineering. Today, roughly 7,000 team members are organized in small teams of 8 to 12 people, focused on developing outcome-based solutions. One team, as an example, exclusively develops payments solutions, ensuring new features like Apple Pay or Google Pay can be deployed across the business.
As consumers spend more time online on the hunt for a new home insurance policy, they are confronted with questions about the age of their roof and other key details they need to input before getting a quote for a plan. Generative artificial intelligence advancements, Jeevanjee says, can help make the process easier and smoother for consumers (and ideally, Allstate hopes, increase business opportunities).
Internally, Allstate attempted to develop a tool to train AI models to automate this process, but determined it had insufficient imaging data. Instead, Allstate is working with a third-party AI startup that takes satellite images of home roofs, which Allstate can use to determine a home policy rate without getting that information from a policyholder. This feature is being used in several states including Illinois and Tennessee.
Other applications of generative AI include a copilot tool custom built by Allstate that initially was tested by 50 employees, but is now used by all 14,000 who investigate insurance claims. Generative AI helps ensure that the 50,000 messages sent to customers each day remove unnecessary insurance jargon and feature more empathetic language. Jeevanjee says the new tool has helped improve a metric Allstate tracks to assess a customer's experience.
Another generative AI pilot currently underway is a so-called “customer engagement sidekick,” which listens into a conversation with a customer and guides sales agents. If the customer mentions their basement has flooded a few times, and the agent forgets to follow up on that detail, the generative AI tool will share a reminder to ask the customer about flood protection. Jeevanjee says 30 employees use this tool today and the plan is to roll it out to all 20,000 licensed sales representatives and call center employees.
Increasingly, while AI is doing a lot of work for employees, the decisions are still being made by humans. But that too may change.
“We’re being a little bit cautious right now,” says Jeevanjee. “At some point, we’ll get comfortable letting AI actually have the conversation and have the human in the background.”
John Kell
Send thoughts or suggestions to CIO Intelligence here.
