WSO2 Applies AI to API Management

AI Accelerating AIOps with the Mainframe

Content By Devops .com

WSO2 today unfurled a Choreo cloud platform through which DevOps teams can create and manage application programming interfaces (APIs) more efficiently by leveraging machine learning algorithms and other forms of artificial intelligence (AI).

Eric Newcomer, WSO2 CTO, said the integration platform-as-a-service (iPaaS) eliminates all the tasks associated with setting up a platform for and managing services based on APIs.

Choreo, available in beta, employs AI to learn historical activities and performance behaviors in order to anticipate the needs of a developer. As developers build an API, the platform will provide performance feedback, identify anomalies, map data and even complete code, said Newcomer.

The Choreo platform also comes with built-in observability tools for monitoring API usage and troubleshooting APIs, added Newcomer.

In addition, Choreo provides access to the open-source Ballerina programming language the company developed to enable developers to create APIs using both low-code and procedural code as they see fit, noted Newcomer. The low-code diagram that developers create generates source code that developers can then run wherever they want, said Newcomer. Developers can close the entire repository Choreo provides, Newcomer added.

Developers can also trigger low-code integration logic to control hundreds of software-as-a-service (SaaS) APIs via pre-built connectors and templates provided by WSO2 or ones they create themselves.

WSO2 is also making it possible to deploy APIs via a single click in Kubernetes environments, which eliminates the need to manage cumbersome YAML files. That capability is based on technology that WSO2 gained via the acquisition of Performer that the company revealed today.

Finally, WSO2 is making available an API marketplace through which organizations can share APIs, event streams, data sources and templates.

With the rise of microservices-based applications, API management has become a lot more challenging. Each additional microservice adds an API that needs to be managed and maintained. As more APIs are employed, it becomes increasingly difficult for DevOps team to keep track of the dependencies that exist between all those APIs without the aid of machine learning algorithms.

It’s not likely machine learning algorithms will replace the need for DevOps professionals any time soon. However, it’s becoming apparent that many of the manual tasks that DevOps teams routinely perform today are becoming automated by machine learning algorithms. In other cases. DevOps teams are gaining new capabilities that would have either been impossible for them to do on their own or would simply have taken too much time to complete.

Of course, it takes time for any AI platform to learn the DevOps environment. IT teams need to invest a significant amount of time training an AI model to recognize relevant issues in much the same way they might train any new member of a DevOps team. The only difference is, once a machine learning algorithm learns something, they never forget it. Nor do they take a day off or announce one day they are leaving the company because they got a better offer somewhere else.

It may take a while before DevOps teams are able to strike the right balance between man and machine, but one thing that is certain is that AI in DevOps environments is here to stay.

Leave a Reply

Your email address will not be published. Required fields are marked *