Cloud Engineering leader Pulumi today announced WhyLabs, the AI observability platform, chose the Pulumi Cloud Engineering Platform to build, deploy, and manage its cloud applications and infrastructure and ready them to adopt multi-cloud deployments. With Pulumi, WhyLabs can continuously and reliably ship new features faster than before for improved time-to-market and enable the company to adopt cloud engineering best practices out-of-the-box, including building infrastructure as code with general-purpose languages (TypeScript, Python, Go, and C#), versioning and reviewing infrastructure code and deploying through a CI/CD pipeline.
WhyLabs is focused on helping companies adopt AI. The WhyLabs observability platform monitors data pipelines and machine learning applications for data-quality regressions, data drift and model performance degradation. This provides transparency for teams building and operating data and AI-enabled applications and reduces manual troubleshooting when a model no longer fits the underlying data used by an application. The company was founded in December 2019 and emerged from stealth in 2020, after raising a $4 million seed round. It raised its $10 million Series A round nearly a year later.
Moving to Pulumi allowed the WhyLabs engineering teams to apply standard software engineering practices and tools uniformly across infrastructure management, application development, and security to manage the complexity of delivering and managing modern cloud applications. Key advantages for WhyLabs included:
According to Andy Dang, co-founder and lead engineer for WhyLabs, he first learned about Pulumi from a colleague at AI2. "I was extremely impressed by the multi-language experience as well as the various innovations, such as its best-in-class support for Kubernetes," said Dang. "Those qualities, as well as support for multiple cloud providers was the deciding factor in choosing Pulumi. And this decision is serving us extremely well."
Prior to adopting Pulumi, WhyLabs tried out AWS CloudFormation and the AWS Cloud Development Kit (CDK) to manage their infrastructure. However, these tools frequently made infrastructure management more complex and inefficient for developers, which stifled WhyLabs' ability to innovate quickly as a young startup. For example, developers had trouble using CloudFormation's domain-specific language (DSL), which made it difficult and time-consuming to model infrastructure. Developers found the DSL limiting since it didn't support standard language features like conditions, loops, and constants. It also didn't enable standard software principles such as creating modular and reusable infrastructure. In addition, these tools only supported AWS when WhyLabs needed the flexibility to go multi-cloud.
"Shifting infrastructure left was critical to accelerating development velocity at WhyLabs," said Aaron Kao, vice president of Marketing for Pulumi. "The engineers at WhyLabs picked Pulumi because it is designed to easily fit into existing software development workflows and empower developers to manage infrastructure as part of the application. With Pulumi, all infrastructure changes followed a code review process, ran through automated tests, and maintained compliance through policy-as-code. Infrastructure bugs and deployment issues were detected faster which led to quicker iterations and faster velocity for WhyLabs."
WhyLabs uses modern cloud architectures for its AI applications running on AWS. They run containers with Amazon ECS, serverless with AWS Lambda, and Amazon EMR for large-scale data processing. They also have deployed Amazon CloudFront as their content delivery platform. To read the WhyLabs case study in full, please visit: https://www.pulumi.com/case-studies/whylabs/.
These press releases may also interest you