Worked on enhancing Vertex AI compatibility for CrewAI samples within the a2aproject/a2a-samples and a2aproject/A2A repositories, focusing on robust environment configuration and API integration using Python and Markdown. Delivered end-to-end Vertex AI readiness by implementing environment variable checks, updating agent image generation models, and introducing configurable client timeouts to improve runtime reliability. Updated documentation to streamline onboarding and clarify setup and troubleshooting steps. Addressed potential API key authentication issues to reduce deployment risk in Vertex AI workflows. The work emphasized error handling and LLM integration, resulting in improved reliability and a smoother user experience for new and existing users.