The Problem
As an indie developer, I was tired of the endless cycle of "what should I build next?" I'd spend weeks browsing Reddit threads, analyzing competitors, and hunting for user pain points instead of actually building.
The traditional approach felt broken:
Most "pain points" on forums are just people venting, not real market opportunities By the time you finish market research, your motivation is gone Building in isolation leads to products nobody wants
The Solution
Instead of developers hunting for ideas, why not flip it? Let AI monitor where users naturally complain (Reddit, Twitter, niche communities) and transform those pain points into actionable product opportunities.
How it works:
AI monitors social platforms for genuine user frustrations Sentiment analysis + frequency scoring identifies promising opportunities Community of developers votes and refines each idea Daily curated inspiration delivered with market context and technical suggestions
Technical Implementation
Stack:
Frontend: Next.js 15.3.3 + React 19 + TypeScript Backend: Cloudflare Workers + Hono framework Hosting: Cloudflare Pages with OpenNext.js UI: shadcn/ui + Radix primitives + Tailwind CSS 4.1.1 Internationalization: next-intl for English/Chinese support
Why Cloudflare's edge stack:
Global performance: Edge Workers eliminate cold starts Cost efficiency: Scales to zero, pay per request Simplicity: Single platform for static assets, APIs, and compute
Interesting technical choices:
Using OpenNext.js to run Next.js 15 on Cloudflare Workers (still experimental but works great) Built custom AI pipeline for social listening and sentiment analysis Leveraging Workers KV for caching API responses globally Full TypeScript across frontend/backend with shared types
Current Status
1,000+ developers subscribed to daily inspirations 10+ product ideas curated and archived Community discussions with voting and feedback WeChat group for Chinese developers (because why not?)
Revenue: Currently free (exploring premium features like AI market validation) What I Learned
Cloudflare Workers are underrated for full-stack apps. The DX is excellent once you get the hang of it, and the global performance is unmatched. Community-driven curation works. Having developers vote on ideas dramatically improves quality vs pure AI selection.
Internationalization from day one was the right call. ~40% of users are Chinese developers. Next.js 15 + React 19 on edge workers feels like the future, even if slightly bleeding edge.
Challenges
Social API limits: Had to get creative with data collection methods Edge runtime constraints: Some Node.js packages don't work in Workers Balancing automation vs curation: Pure AI gives too many false positives
What's Next
Working on:
AI-powered market validation for submitted ideas Collaborative spaces for community projects Success tracking (inspiration → shipped product) More data sources beyond social platforms
Questions for HN
Fellow indie developers: What's your biggest challenge in finding product ideas? Technical: Anyone else running Next.js 15 on Cloudflare Workers? How's your experience? AI/ML folks: Best approaches for sentiment analysis on social data at scale? Community building: How do you maintain quality in user-generated content platforms?
Try it out: builtwhat.app
Happy to answer questions about the technical implementation, community growth, or anything else!
Built this mostly for myself and other indie developers who were stuck in analysis paralysis. Turns out I wasn't alone.