Conversational AI platform modernized infrastructure with microservices architecture and automated deployments

Infrastructure & DevOps Modernization
AWS Foundations

Company

Share

Modernizing their infrastructure helped Valyant AI scale their services and streamline their infrastructure, while improving their overall efficiency and performance.

Valyant AI is a conversational AI platform transforming customer service in the Quick Serve Restaurant (QSR) industry through technologies such as AI enabled drive-thru, mobile, kiosk, and text message ordering. Founded in Denver, Colorado, in 2017, Valyant AI’s mission is to accelerate customer ordering, reduce overhead and increase profitability for their QSR customers.


Challenge

On their existing infrastructure, Valyant AI had a manual deployment process that proved inefficient, requiring off-hour deployments to avoid disruptions for their customers. The company was looking for help to improve the process of building secure, high-performing, fault-tolerant and cost-effective cloud-native environments in AWS leveraging microservices and establishing an automated and consistent way to build, test and deploy applications on top of these environments.

Solution

A Caylent engineering pod was engaged to address Valyant AI’s request to improve the deployment process. The pod built a new greenfield deployment environment on Amazon ECS which was compliant with AWS best practices. This included using Terraform to provision infrastructure, using AWS CodePipeline to automate deployments, and using AWS CodeDeploy to distribute traffic.

The pod began improving Valyant AI’s infrastructure pipeline process by updating Terraform Cloud CI/CD pipelines with a CodeDeploy blue-green deployment process. This required updating the Terraform Cloud configuration to use CodeDeploy, a fully managed service that automates deployments to Amazon EC2, AWS Fargate and on-prem instances. An internal load balancer was removed, replaced by CodeDeploy’s automated traffic distribution between Valyant AI’s environments.

Valyant AI’s compute platform was also migrated from Amazon Elastic Container Service (ECS) on EC2 to ECS on Fargate, a serverless compute service that allows running containers without having to manage servers. This simplified scaling the deployment environment and reduced costs while also reducing latency within ECS.


Results

The changes made by Caylent resulted in a significant improvement to the deployment process. The new process is more reliable, efficient, and scalable. The new environment was built with clear and repeatable processes where testing could be isolated. It has also made it easier to deploy new changes and to troubleshoot problems.

Valyant AI was able to improve their engineering team's working conditions by making deployments easier and more efficient. By leveraging microservices architecture with ECS, Fargate, and implementing blue-green deployment leveraging automated CI/CD pipelines, ValyantAI was able to:

  • Reduce manual tasks
  • Accelerate time to deployments
  • Ensure critical deployments were successful and followed best practices

Additionally, new workloads have the necessary resilience to meet Valyant AI’s needs. Modernizing their infrastructure helped ValyantAI scale their services and streamline their infrastructure, while improving their overall efficiency and performance.

Company

Share

Related Services

Related Case Studies

Availity Logo

Availity

Healthcare information network powers data upcycling with Kubernetes on AWS

Read more
SigFig Logo

SigFig

Financial services technology provider accelerates client onboarding and increases resilience by migrating to AWS cloud

Read more