CryptoFund

CryptoFund: Crowdfund on Ethereum with Gemini AI—predict success, seize market trends.

Comment

CryptoFund

Inspiration

The spark for CryptoFund came from the growing wave of Web3 innovation and my passion for making decentralized systems accessible. Crowdfunding platforms often lack transparency—donors wonder where their money goes, and creators struggle to stand out. I saw an opportunity to blend Ethereum’s trustless blockchain with AI’s predictive power, inspired by the GDG Solution Challenge’s call to solve real-world problems. I wanted to empower creators globally while harnessing Gemini AI to make funding smarter, especially as crypto adoption soars.

What it does

CryptoFund is a decentralized crowdfunding platform on Ethereum, supercharged by the Gemini AI Model. Creators launch campaigns with clear, on-chain transactions, while donors track every ETH via Etherscan. The AI predicts campaign success, forecasts ETH market trends using real-time Finnhub data, and offers tailored advice—like when to promote based on price spikes. It learns from itself, improving predictions and delivering insights across finance, user behavior, and more, all seamlessly integrated into a sleek Next.js app.

How we built it

We built CryptoFund with a lean, modern stack:

  • Frontend: Next.js 15.1.0 and Tailwind CSS for a responsive, user-friendly interface.
  • Backend: Node.js and MongoDB for campaign, user, and transaction data.
  • Blockchain: Ethereum smart contracts for secure donations, with Finnhub feeding live ETH/USD prices.
  • AI: Integrated Google’s Gemini Model for self-learning predictions and insights, trained on synthetic campaign data (1,000+ entries) and market trends.
  • Authentication: NextAuth.js for secure user and admin access.
  • Deployment: Hosted on Vercel at contrichain.vercel.app.

We started by designing smart contracts, then built API routes to fetch Finnhub data and query Gemini AI. The frontend came last, with real-time dashboards and AI-driven advice woven throughout. Flagsmith toggled features for rapid iteration during the hackathon.

Challenges we ran into

The hackathon wasn’t without hurdles:

  • AI Integration: Getting Gemini’s self-learning model to process real-time Finnhub data reliably took tweaking—early predictions were noisy.
  • Blockchain Sync: Ensuring Ethereum transactions reflected instantly in the UI was tricky with network delays.
  • Data Pipeline: Merging campaign stats and market data into MongoDB for AI training hit scaling issues; we optimized queries under time pressure.
  • Time Crunch: Balancing a polished frontend with backend complexity in 48 hours meant late-night debugging sessions!

We tackled these with teamwork, quick pivots, and lots of coffee—learning to prioritize what mattered most.

Accomplishments that we're proud of

  • Live Deployment: Launched a fully functional app at contrichain.vercel.app during the hackathon.
  • AI Innovation: Built a self-learning Gemini AI that predicts campaign outcomes and ETH trends with growing accuracy.
  • Transparency: Delivered end-to-end blockchain tracking, making every donation verifiable.
  • User Experience: Crafted an intuitive UI that hides Web3 complexity, earning early tester praise.
  • Hackathon Spirit: Pushed boundaries by blending Web3, AI, and real-time data in a tight deadline.

What we learned

This journey taught us heaps:

  • Web3 Nuances: Smart contract optimization and gas fees are an art—we got better at both.
  • AI Power: Gemini’s self-learning opened our eyes to dynamic predictions beyond static models.
  • Real-Time Data: Finnhub’s API showed us how live markets can shape user decisions.
  • Teamwork: Crunch-time collaboration honed our ability to divide, conquer, and iterate fast.
  • User Focus: Feedback loops revealed how much simplicity matters in Web3 apps.

What's next for CryptoFund

CryptoFund’s just getting started! We’re planning to:

  • Enhance AI: Train Gemini on real user data for even sharper predictions and multi-asset market insights.
  • Expand Chains: Support Polygon or Solana for lower fees and broader reach.
  • Community Features: Add forums and rewards to engage backers more deeply.
  • Mobile App: Build a native app for on-the-go funding.
  • Global Impact: Partner with nonprofits to fund social good projects via CryptoFund.

We’re excited to keep pushing Web3 and AI forward—join us at contrichain.vercel.app!

Built With

Share this project:

Updates