Cloud-Based Agility in Responding to Disasters

When we talk about enterprise agility in the cloud, we’re often talking about achieving it through DevOps or other Agile software delivery approaches — about the ability to quickly provision infrastructure, load it up with deployed software, receive fast feedback from users, and scale up and scale down instantaneously. But software agility is only one type of enterprise agility that the cloud can enable.

Starting on August 25, 2017, Hurricane Harvey hit the coast of Texas, pouring almost 52 inches of rain on the Houston area and causing $180 billion of damage[1]. The American Red Cross jumped in by deploying hundreds of volunteers to open shelters and provide food, comfort, and support to people frantically looking for help. Ultimately, the Red Cross worked alongside partners to provide 414,000 overnight stays in emergency shelters, serve 4.5 million meals and snacks, and distribute 1.6 million relief items. Thanks to the generosity of the public, the Red Cross was also able to provide $229 million in direct financial assistance to some 573,000 of the most severely affected households in the first two months[2]. Harvey wasn’t the only large disaster to hit the United States in the fall of 2017. In fact, in just the 45 days after Harvey made landfall in Texas, the Red Cross responded to five additional large and complex disasters, including back-to-back hurricanes — Irma, Maria and Nate.

Because it must respond quickly to unpredictable disasters, the Red Cross mission always requires agility. But, in the case of Harvey, even its adaptive abilities were challenged: the scale of the damage and the amount of assistance available combined to cause a flood of phone calls that quickly overwhelmed the Red Cross call center. As a result, the Red Cross engaged the services of an AWS Partner Network (APN) partner, Voice Foundry, an expert in the implementation of Amazon Connect. Amazon Connect is a self-service, cloud-based contact center service that’s based on the same technology used globally by Amazon customer service associates. Within 48 hours, a new call center was in operation, and Amazon employees were taking calls for three of the hurricanes and geographical areas. This helped supplement the usual Red Cross volunteer call centers. When the peak volume of calls subsided two weeks later, the Amazon call center was de-provisioned.

This triumph of enterprise agility was made possible by the Red Cross culture of rapid response, its ability to mobilize and focus efforts at the time of a disaster, and the deep experience of an APN partner who knew just how to respond. As with Agile software delivery, this victory was also achieved through the actions of a small, cross-functional team that worked across organizational silos.

In this case, the team consisted of three Voice Foundry specialists who were supported by twice-daily calls that included leaders of the Red Cross, Voice Foundry, and Amazon. The team was assembled overnight and, early the next morning, it began analyzing and re-working the call routing rules while training the Amazon operators to handle what were often highly emotional phone calls. Their focus was on creating a minimal viable solution that could begin adding value quickly. And, as in any good DevOps process with continuous integration at its core, the team rapidly integrated the various volunteer call centers into a single call routing system. In the 48 hours that it took to provision the new phone lines, the team had set up an entirely new call center, established the routing rules, and trained the new operators — a process that would normally have taken 4–5 months for a call center of equivalent size.

To me, one of the most interesting things about this event was not just the speed at which the call center was assembled, but the speed with which it was disassembled as well. When it was needed, it appeared; when it had served its purpose, it was gone. This is precisely the kind of elasticity that makes cloud infrastructure such a powerful driver of agility — but how interesting that this elasticity could exist, and be deployed, for an entire business function! This, I think, is typical of today’s cloud, where even pre-trained machine learning models can be made available to support new ways of doing business within moments.

When I think about true organizational agility, I also think about Hess, which found itself with a need to rapidly divest a number of its businesses. By migrating to the cloud and the AWS platform, the energy company was able to prepare its IT assets to move to the acquiring companies in just six months.

Or the Centers for Medicare and Medicaid (CMS), which developed three new products on AWS after the original challenging launch of healthcare.gov, with no further problems in scalability or responsiveness.

Or, perhaps, the Louisiana Department of Corrections, which faced the challenge of getting inmates heavily controlled access to the Internet, so they could prepare to get jobs when released from prison. The solution here centered on virtual desktops provided by Amazon Workspaces.

Each of these cases illustrates how organizations can achieve new kinds of agility under demanding circumstances by using the cloud. But the story of how the Red Cross rapidly scaled up and scaled down its disaster relief efforts in response to the unique circumstances surrounding Hurricane Harvey is an unmatched, dramatic example of using the cloud to achieve organizational agility.

[1] Kimberly Amadeo in The Balance, “Hurricane Harvey Facts, Damage, Costs,” https://www.thebalance.com/hurricane-harvey-facts-damage-costs-4150087, September 30, 2017.

[2] American Red Cross blog, “Hurricane Harvey Response: At the 2 Month Mark,” http://www.redcross.org/news/article/local/texas/gulf-coast/Hurricane-Harvey-Response-At-2-Month-Mark, November 2, 2017.