Onpost

About the Project

ONPOST is a forum-first, AI-assisted marketplace where real community posts evolve into structured markets and actionable price insights. Instead of forcing users into rigid listings, ONPOST lets people post freely—and only turns activity into analytics when the data is real and sufficient.

Inspiration

Most online marketplaces feel artificial at the start: empty listings, fake prices, and little real demand.
Meanwhile, many real trades already happen organically inside forums, chats, and comment sections—but that data is never structured or analyzed.

I was inspired by a simple question:

What if markets didn’t start with listings, but emerged naturally from real conversations and trade posts?

ONPOST explores that idea by treating posts as the source of truth and letting markets, prices, and insights form only when community activity grows.

What I Learned

This project taught me several important lessons:

• Data maturity matters: Analytics are meaningless without enough real data. That’s why ONPOST uses thresholds (e.g. 50 valid trades) before unlocking insights.
• AI doesn’t need to run all the time: On-demand AI, triggered only by user intent, is more efficient and more trustworthy.
• UX clarity beats complexity: Separating Live Feed (free posts) from Market Analytics (structured threads) made the product much easier to understand.
• Not everything needs a database: By carefully using extendedData, it’s possible to build derived intelligence without external storage.

How I Built It

ONPOST is built using a forum-first architecture:

• Foru.ms as the core backend:
  • Threads, posts, users, private messages
  • extendedData used to store lightweight trade and analytics metadata
• Next.js + React + Tailwind for the frontend:
  • Live Feed for free posting
  • Market pages for analytics and insights
• AI Market Assistant:
  • Lives in the sidebar
  • Activates only when users ask questions
  • Reads real posts via API
  • Computes price stats (min / median / max)
  • Uses AI only to summarize structured data

Incremental Analytics Logic

Instead of scanning everything repeatedly, ONPOST processes data incrementally using cursors:

New posts → Trade detection → Validation → Count threshold → Analytics unlock

Built With

• foru.ms
• next.js
• react
• tailwind
• typescript