Вештачка интелигенција

CircleCI Expands CI/CD Platform for AI Integration in DevOps Workflows

Summary

CircleCI, a leading provider of continuous integration and continuous delivery (CI/CD) solutions, announced this week that it is expanding the capabilities of its platform to enable the seamless integration of artificial intelligence (AI) models in DevOps workflows. In addition to […]

CircleCI Expands CI/CD Platform for AI Integration in DevOps Workflows

CircleCI, a leading provider of continuous integration and continuous delivery (CI/CD) solutions, announced this week that it is expanding the capabilities of its platform to enable the seamless integration of artificial intelligence (AI) models in DevOps workflows. In addition to providing access to the latest generation of Graphics Processing Units (GPUs) from NVIDIA via Amazon Web Services (AWS) cloud, CircleCI has added incoming webhooks to access AI model curation services from providers such as Hugging Face, as well as integration with LangSmith, a debugging tool for generative AI applications, and Amazon SageMaker, a service for building AI applications.

The CEO of CircleCI, Jim Rose, stated that while there is a lot of enthusiasm for incorporating AI models into applications, the processes used are still immature, especially when it comes to automating workflows that involve testing probabilistic AI models. Rose noted that most AI models are built by small teams of data scientists who create a software artifact that needs to be integrated within the DevOps workflow like any other artifact. The challenge lies in the fact that most data science teams have not yet defined a set of workflows to automate the delivery of these artifacts as part of a broader DevOps workflow, he added.

DevOps teams also need to adapt their approach to version-based application management in order to trigger flows for extracting AI software artifacts that exist outside traditional software repositories. For example, the incoming webhooks provided by CircleCI now enable automatic pipeline creation when an AI model hosted on Hugging Face changes.

It is still early days for implementing AI models in production environments, but there is no doubt that generative AI will have a significant impact on software development. AI models are a different class of software artifacts that are continuously retrained instead of being periodically updated. As such, DevOps teams need to keep track of every moment when an AI model is retrained to ensure application updates.

At the same time, generative AI will also accelerate the pace of creating and implementing other software artifacts. Many manual tasks that currently slow down the speed of application development and deployment will be eliminated. This does not mean that there won’t be a need for software engineers, but it does mean that the role they play in software development and implementation will rapidly change. DevOps teams must assess how generative AI will impact the tasks they manage and how the overall Software Development Lifecycle (SDLC) needs to evolve.

As always, each organization will need to decide for itself how to best achieve these goals depending on their AI use cases, but the changes brought by generative AI are now inevitable. The longer the adaptation takes, the harder it will be to overcome the cultural and technical challenges that will be encountered along the way.

FAQ:

Q: How is CircleCI expanding its CI/CD platform?
A: CircleCI is expanding its CI/CD platform to enable the seamless integration of AI models in DevOps workflows.

Q: What services have been added to the CircleCI platform?
A: The CircleCI platform has added services to access AI models from curation services such as Hugging Face and integration with LangSmith and Amazon SageMaker.

Q: What impact will AI models have on software development?
A: AI models will have a significant impact on software development, accelerating the pace of creating and implementing software artifacts.

Q: How are DevOps teams adapting their approach to introducing AI models?
A: DevOps teams are adapting their approach by changing how they manage applications and defining new workflows for automating the delivery of artifacts.

Q: What should organizations do to achieve their goals in using AI?
A: Each organization should independently decide how to achieve their goals in using AI based on their specific use cases, but the changes that AI brings are inevitable and require adaptation.

Source: circleci.com