Optimizing global networking operations using AWS Cloud WAN with TFS

TFS is the brand for finance and related products for Toyota in the United States. As one of the largest auto financial services companies in the world, TFS provides financing, leasing, and insurance offerings for millions of Toyota’s customers. Established in 1982, TFS has grown from an 8-member team to a company with more than 3,800 team members and over USD $150 billion in managed assets.

As part of its cloud transformation, TFS wanted to improve its reliability and availability to support its increasing needs and growing customer base. The company also wanted to boost developer productivity through self-service tools and a better overall experience. That’s why TFS realized the need to use a cloud-centered networking platform to support its business growth.

Using AWS services, TFS began rebuilding its environment across multiple AWS Regions. The company had already been using AWS for some workflows and wanted to consolidate its infrastructure. “We chose AWS for the depth of the service offerings and the availability of talent in the market that comes with related skill sets,” says Jim Stanley, national manager of cloud engineering at TFS. “Our success in this transformation journey requires that we can hire talent with the right skills.”

TFS worked with the AWS team to do technical deep dives on the services that the company would adopt. The company planned to replicate its existing network segmentations by using AWS Cloud WAN, a service for building, managing, and monitoring global wide-area networks.

The company started simple, implementing AWS Cloud WAN for two virtual private clouds (VPCs), and then tested more services and infrastructure improvements. Next, TFS began scaling out its firewalls by using Gateway Load Balancer, a service for deploying, scaling, and running third-party virtual appliances. To manage IP addresses in a dynamic networking environment, the company uses Amazon VPC IP Address Manager (Amazon VPC IPAM), a feature that makes it simple to plan, track, and monitor IP addresses for AWS workloads.

“We’re using the combination of AWS Cloud WAN, Amazon VPC IPAM, and Gateway Load Balancer to achieve the resiliency, speed, and scalability that we want,” says Chris Gama, senior cloud engineer at TFS.

TFS has maximized its scalability on a global scale by using AWS services. “We can scale our perimeter firewalls more simply by using AWS networking services,” says Gama. To scale out, the company deploys a new firewall and adds it to the group. TFS has also set up a multi-Region network by using AWS Global Infrastructure, so failover is seamless and security is stronger. “This is a benefit of using first-party solutions, where the connection between services is so natural,” says Stanley. “It simplifies processes and reduces administrative overhead.”

TFS uses Amazon VPC IPAM in its account-vending applications. “Using Amazon VPC IPAM was an easy choice,” says Stanley. “It was simple for us to set up a few pools and update our code. In addition, it gave us deeper visibility into what IP addresses are being used and where.” This way, TFS is achieving its goal of being fast and agile.

The company has also improved developer productivity by automating the account-vending process, which was time consuming and involved finding available networks, planning routing changes, and creating VPCs. That process required working across multiple teams and took up to 2 weeks. Now, TFS can determine the necessary inputs and call an API to activate the automatic account-vending system in 1 hour.

“Our engineers have more time now to focus on higher-order functions, working on more important projects,” says Stanley. TFS’s developers no longer need to wait for the infrastructure team to make changes. This means that they can build and release applications quickly to improve the experience for consumers and dealers.

On AWS, TFS minimized downtime and removed the need to manually manage its infrastructure. “A welcome result of the migration is that we don’t need to think about software versions, apply patches, or manage planned upgrades,” says Stanley. “A couple of times, it felt like we were missing some complexity somewhere. The setup was painless, the migration was straightforward, and the solution has been refreshingly low maintenance since going live.” The company can set up a global network in minutes and get back to more important tasks that improve the experience for the users of its services.

All TFS’s new infrastructure runs on AWS networking services. The company migrated about 60 percent of its existing systems after 1 month and completed the migration in another month. Reflecting on the experience, TFS regrets not undertaking the migration sooner. “A big takeaway from this experience is that cloud networking doesn’t need to be difficult,” says Stanley. “AWS does a good job of abstracting a lot of the complexity.” Now, TFS can deliver products faster, with total infrastructure provisioning and application onboarding time reduced from 10 weeks to 2 hours.

The company continues to look for ways to optimize operations and innovate, including exploring cutting-edge technologies to streamline processes. “We are considering some generative AI projects that can improve the capabilities that we’ve built using AWS,” says Stanley. “All our modernization initiatives build on the foundation that we’ve set on AWS.”

TFS has transformed its infrastructure to support its growth as a cloud-centered company. “By using AWS and focusing on native services, we’re removing the unnecessary complexity that robs our engineers of capacity and focus,” says Stanley. “Our experience along this journey of prioritizing AWS services to streamline integration, simplify administration, and accelerate engineering onboarding has been really beneficial.”