PikaEdge

Interactive map-based platform for discovering real-time arbitrage opportunities across Malaysia, Singapore, and Japan. Track trending products and find profitable cross-border deals.

Comment

Inspiration

I was intrigued when my brother and friends got into trading cards. I was surprised to discover the high turnover value for each card—often better returns than crypto and far more stable!

What it does

PikaEdge scours the internet for hard-to-find deals and updates on rare cards across Malaysia, Singapore, and Japan. It surfaces arbitrage opportunities by tracking price differences, trending items, and real-time market activity—helping traders discover profitable flips before they hit the mainstream.

How I built it

I used Cursor as our main development tool, building the frontend with Next.js and shadcn/ui, and Convex for the backend. I was genuinely surprised by how powerful Claude Opus 4.5 is—combined with a few screenshots from Mobbin, it produced exceptional results that matched my vision perfectly!

Challenges I ran into

Figuring out reliable data sources was the biggest hurdle. eBay requires a 24-hour API review process, most platforms have limited public APIs, and X.com's API is behind a paywall. We had to pivot to web crawling and creative data aggregation to get the real-time information we needed.

Accomplishments that I'm proud of

I got the entire backend working in just a few hours, complete with functioning Convex functions and cron jobs. Cursor's code review feature caught several bugs before we even committed, saving tons of debugging time. The Convex MCP integration was also invaluable for setting up the database schema and streamlining deployment.

What I learned

Backend development is still challenging, but Convex made it so much easier—no more wrestling with SQL queries! Real-time data synchronization and serverless functions just work out of the box.

What's next for PikaEdge

  • Expand data sources to cover more marketplaces and regions
  • Integrate social sentiment analysis from X.com, Reddit, and trading communities
  • Build predictive alerts for upcoming valuable cards before they hit mainstream markets
  • Add price history tracking and trend forecasting for better trading decisions
  • Use LLM to analyze user sentiments on the cards (I hear from friends all the time!)

Built With

  • claude
  • clerk
  • convex
  • cursor
  • firecrawl
  • mobbin
  • next.js
  • shadcn
  • vercel
Share this project:

Updates