Can LLMs generate Enterprise Quality Code

This talk explores the creation of reliable, maintainable, and secure code for enterprise applications using modern AI agents. The speaker discusses benchmark testing conducted by Sonar on 27 of the latest large language models, assessing their task completion and code quality. The session highlights the distinctions among these models, revealing that some can generate significantly more code issues than others. Additionally, it covers the integration of AI agents with deterministic static analysis to maintain enterprise-level quality while preserving the productivity benefits of AI.