How I Use AI for Product Work (elezea.com)

A product manager shares insights on utilizing large language models (LLMs) as a collaborative "sparring partner" in their daily product work. Rather than relying on AI to generate core outputs, they emphasize using it to refine ideas, challenge weak reasoning, and fill in knowledge gaps. This approach encourages critical thinking, making AI a tool for enhancing clarity and depth in product management rather than simply a ghostwriter. The methodology involves creating a structured system of prompts, personal context files, and reference materials that tailor the AI’s responses to the user's specific style and project requirements. Key elements include opinionated system prompts for different tasks and the ability to connect to real data sources via Model Context Protocol (MCP) servers. This setup allows the AI to provide contextually relevant feedback that aligns with the user's standards, thus fostering better communication and more effective problem-solving. The author concludes that while the AI doesn’t replace their thinking, it significantly augments their ability to develop comprehensive documentation and refine their product strategy.