I built Idea Harvester because I was tired of the endless "what should I build next?" cycle that so many of us face. I'd spend hours scrolling through Reddit, stumbling across interesting problems in random threads, but never had a systematic way to capture and analyze these goldmines of real user pain points. What makes this different from other idea tools? Most idea discovery platforms focus on trends and market data, but I wanted to tap into something more authentic - actual conversations where people are venting about their real problems. Reddit is this incredible treasure trove of unfiltered feedback, but it's scattered across thousands of communities. Here's what I'm most proud of: We filter all noise. We read all posts, and comments. he AI analysis that automatically identifies pain points and suggests potential solutions from Reddit posts The PRD generation and market analysis is both compact and great. The audience management system that lets you define your target communities and keywords The clean, modern interface that makes research actually enjoyable (built with Next.js and shadcn/ui) The webhook-based architecture that keeps everything fast and responsive It's been such a journey getting the AI to properly parse Reddit conversations and extract meaningful insights. The system now categorizes posts by service types, identifies topics, and even suggests solution directions - all while maintaining that authentic voice from the original discussions. Would love to hear what you think! Has anyone else struggled with the "what to build" problem? How do you currently discover new opportunities?