In the era of automation, society has invested significant effort in automating repetitive processes across various sectors to reduce the manufacturing time of many products. However, we have not given similar attention to automating software development, as it involves complex decision-making, contextual understanding, and requires human expertise and coordination. Historically, most organizations followed the waterfall methodology for the Software Development Life Cycle (SDLC), and in the early 21st century, they rapidly adopted agile methodologies with the expectation of delivering more robust and scalable products within a shorter timeframe. However, human involvement has remained central in all these methodologies until the emergence of Agentic AI. Agentic AI has the potential to transform software development in ways that have not been previously explored. In this paper, we propose an agent-driven SDLC framework that adopts an AI-first approach to software development, where human involvement is limited to governance and decision-making. The framework introduces a Central Orchestrator Agent that coordinates with specialized agents responsible for backlog planning, solution architecture, code generation, automated testing, code review, CI/CD, deployment orchestration, and production monitoring with self-healing capabilities. This AI-first approach can significantly reduce human effort and software release time, maintain high code quality through automated validation, and enable rapid incident response through autonomous hotfix generation and rollback mechanisms.