The post details a structured approach to coding at a FAANG company, emphasizing the integration of AI tools in the development process. The author, an AI software engineer with over a decade of experience, outlines the importance of starting with a solid technical design document. This document serves as a proposal that must gain stakeholder approval before moving on to detailed system design, which includes architecture and team integrations. Following this, a design review phase allows senior engineers to critique the initial proposal, ensuring potential issues are addressed early. Once the design passes review, developers engage in sprint planning and backlog development. The actual coding phase uses Test Driven Development (TDD), where AI assists in creating tests before the feature itself is developed. The author notes significant efficiency gains, reporting about a 30% increase in speed from proposal to production deployment. However, this structured process has led to some pushback from the community, with critics arguing that the term "vibecoding" should not apply to such a rigorous methodology. Many commenters contend that true vibe coding should be more fluid and less reliant on extensive planning.