IBM has released a new framework called IBM Concert that uses generative AI and knowledge graphs to help DevOps teams quickly identify and address issues that could impact service availability. This tool, built on the IBM WatsonX platform, collects data from various sources in your IT environment and can be hosted on cloud platforms like IBM or Amazon Web Services or on-premises. Vikram Murali, of IBM Automation, explained how IBM Concert offers a comprehensive view of complex IT environments to help teams manage application risks and compliance. While currently focused on risk and compliance management, IBM plans to expand Concert’s capabilities to include cost management in the future.
IBM has been developing AI models called Granite LLMs ranging from 3B to 34B parameters. These models have been tested on various benchmarks and proven to be effective in synthesizing, correcting, and translating code in multiple programming languages. The 20 billion parameter Granite codebase model powers IBM Watson Code Assistant, enabling natural language generation of SQL code. Different LLMs excel in specific use cases, but organizations may need a combination of models to address various tasks efficiently.
As the pace of application development intensifies, DevOps teams will increasingly rely on AI tools to keep up. However, there may be a delay in implementing these tools, resulting in a backlog of code in pipelines. Machines are expected to take on a larger role in both coding and code management, helping teams stay ahead of the rapid development cycles. IBM’s Concert and Granite LLMs are designed to assist DevOps teams in streamlining processes and improving service availability while adapting to the evolving IT landscape.
Article Source
https://devops.com/ibm-makes-generative-ai-platform-for-devops-available/amp/