We provide access to affordable, high-quality health care to 1.75 million-plus Californians who do not have employer-sponsored health care. We faced unprecedented demand for coverage during the first year of the pandemic. Our on-premises monolithic platform failed several times. And when it didn’t fail, the performance was frequently poor. This led to a series of challenging phone calls with my executive director, in which he asked, “What’s going on and when will you fix it?” At the conclusion of each of these conversations I stated, “If this is happening again next year, you should fire me because it’s my job to make sure this doesn’t happen.”
What made a cloud infrastructure most attractive?
Moving to the cloud lets us rapidly scale up to meet unexpected performance demands on our CalHEERS system, like those we faced that first year of the pandemic. And while our former on-premises system was designed to handle our highest estimated open enrollment season traffic in the fall, it coasted during the rest of the year — not a very good use of a capital-intensive asset.
How does the cloud make work easier for your agency?
Since fall 2020, we’ve never experienced another performance-related issue or outage. We did a lift-and-shift that first year to get to the cloud. Since then, we’ve used cloud-native technology to rearchitect our monolithic solution into microservices. We can now envision and deliver new features extraordinarily fast — in weeks and months versus the months and years required on our legacy platform. Consumer experience is our North Star, so moving to a cloud-based microservices architecture has supported delivery of a slew of consumer experience-related features.
How has the cloud improved your resiliency?
On Jan. 26 this year, the vendor who hosts and manages our consumer call center technology stack reported an unplanned outage, taking our call center completely offline. This was the last week of the open enrollment period, when we receive the year’s highest volume of calls from consumers with questions about eligibility and enrollment.
After confirming the incident was cybersecurity-related and would require weeks or months of recovery time, we called our system integrator, Deloitte, and our cloud infrastructure partner, Amazon Web Services, for help. Seeking something akin to a miracle, I told them, “I need 1,200 call center agents answering inbound phone calls on Monday morning.” They mustered a global team of experts to stand up a brand-new call center with capacity for 30,000 inbound calls per day, complex call handling and recording, and an interactive voice response system in 48 hours. The Deloitte and Amazon team delivered — and Covered California was able to successfully serve residents through the end of our highest-volume open enrollment period.
We couldn’t have done it if we hadn’t already put most of our infrastructure in the cloud and developed capabilities and relationships with key partners that provided us confidence we could move the last remaining elements to the cloud in response to a critical outage.
