AMD has announced the introduction of Agentic AI systems aimed at transforming the software development lifecycle (SDLC) by incorporating AI agents at every stage. This development comes after surpassing previous productivity targets and achieving a 30 percent productivity boost through AI.
Agentic AI is designed to enhance various stages of the SDLC, including code generation, testing automation, bug analysis, and code review. These AI agents can analyze problem reports, identify necessary code modifications, generate unit tests, and prepare architecture summaries for engineer review and approval. The system aims to automate up to 50 percent of the codebase, with some components already seeing over 80 percent of their code generated by AI.
The technology leverages collaborative AI agent swarms, which are capable of discovering solutions independently. Unlike traditional AI systems that mimic human workflows, these agents determine optimal paths to solutions based on defined success criteria, performance, and system constraints. This allows engineers to define the problem and desired outcomes, while the AI agents autonomously navigate the solution space.
The primary users of Agentic AI are software developers and engineering teams looking to increase productivity and efficiency in their workflows. The system enables developers to scale their expertise, allowing multiple AI agents to work in parallel, thus extending the reach of individual productivity.
Work implications: The integration of Agentic AI into software development workflows could significantly augment roles in software engineering by automating repetitive tasks and allowing engineers to focus on more complex problem-solving activities.
Originally reported by AIArticleComputing