HACK-TRACK AI - Never miss another AI hackathon

HACK-TRACK AI is your AI-powered hub for every hackathon. We aggregate events, summarize with AI, track deadlines, and reveal winning patterns — so you never miss an opportunity.

Comment

HACK-TRACK AI — Never Miss a Winning Hackathon Again

HACK-TRACK AI is a production intelligence platform that helps builders discover AI hackathons, understand what actually wins, and ship with data instead of guesswork.

Inspiration

I build products and I kept entering hackathons trying to win.

The problem? I had no idea what judges actually rewarded. I would spend weeks building something polished, only to lose to a simple Chrome extension that solved a small problem in 48 hours.

So I started analyzing what actually won. After studying 50+ past hackathon winners, clear patterns emerged:

  • Chrome Extension + AI combos had ~70% win rate in micro-SaaS categories
  • Live demos outperformed static screenshots by a large margin
  • Shipping fast beat over-polishing
  • Next.js + Railway allowed me to go from code to live URL in hours

The insight was simple: most builders are flying blind. We guess what judges want instead of knowing.

That’s why I built HACK-TRACK AI ** — the tool I wished I had. A single place to discover opportunities **and understand what wins before I start building.

What It Does

HACK-TRACK AI helps builders in three key ways:

1. Discover

  • Curated list of AI hackathons from Devpost, MLH, AngelList, and partner networks
  • Filter by prize pool, deadline, difficulty, location, and tech stack
  • Grid, List, and Calendar views
  • One-click Google Calendar export

2. Learn

  • Analyze real winning patterns from 50+ past hackathons
  • See what judges reward: tech stacks, team size, product type, and execution style
  • Insights organized by category and difficulty level

3. Ship Smarter

  • Understand which opportunities get attention
  • Make data-driven decisions about where to participate
  • Stop wasting time on the wrong hackathons

Built for builders, product people, and makers who want to stop guessing and start winning.

How I Built It

Frontend

Next.js 14 + TypeScript + Tailwind CSS

  • Responsive Grid, List, and Calendar views
  • Full dark mode with proper accessibility
  • Smooth animations and mobile-first design
  • Dynamic filtering and Google Calendar export

Backend & Database

Supabase (PostgreSQL + pg_cron)

  • 21+ hackathons aggregated from multiple sources
  • Automated refresh every 10 days using pg_cron
  • Auto status updates (open → closing_soon → closed)
  • Row-level security enabled

Analytics

Novus.ai

  • Automatic tracking of page views, filters, and searches
  • Daily Slack digests with engagement metrics
  • Real-time visitor alerts

Deployment

Railway (Free tier)
Live: https://hacktracker-production.up.railway.app/

Challenges I Ran Into

Dark Mode Implementation

Initial dark mode only worked on the header. Text became invisible on dark backgrounds.
Fix: Added darkMode: 'class' in Tailwind config and systematically updated 30+ components with proper dark variants.

Environment Variables in Next.js

NEXT_PUBLIC_* variables are baked in at build time. Setting them in Railway didn’t work until I triggered a rebuild.

Analytics Setup

Novus.ai showed zero events for the first 24 hours because the environment hadn’t been rebuilt after installation.

Lesson: Always test end-to-end after every integration.

Accomplishments I'm Proud Of

  • Shipped a real, usable product — not a demo. All core features (filters, views, calendar export, dark mode) work in production.
  • Aggregated 21+ hackathons with $200K+ in tracked prizes from Devpost, MLH, and AngelList.
  • Revealed clear winning patterns from 50+ past projects.
  • Achieved zero infrastructure cost using Supabase and Railway free tiers.
  • Built a clean, accessible, and fully responsive experience with thoughtful details.

What I Learned

  • Building for yourself is incredibly powerful — I knew exactly who I was building for because I am the user.
  • Depth beats breadth. Cutting AI summaries and auto-scraping allowed me to deliver real value faster.
  • Dark mode forces you to be intentional about every design decision.
  • Real analytics (via Novus) changed how I prioritized features — builders spent the most time in the “What Wins” section.
  • No existing platform shows winning patterns. That gap made HACK-TRACK AI feel necessary.

What's Next for HACK-TRACK AI

Phase Focus Timeline
Phase 2 Real-time scraping from Devpost & MLH Next 4 weeks
Phase 3 Email deadline alerts Q3
Phase 4 Bookmarking + Team matching Q4
Phase 5 Mobile app (React Native) 2027
Phase 6 Pro tier + Monetization 2027

The Vision

I built HACK-TRACK AI to give builders an unfair advantage.

Not guessing what judges want.
Knowing it.

Not searching manually.
Discovering instantly.

Not shipping blindly.
Shipping to win.

If I’m entering another hackathon, this is where I start. And if other builders use it to win faster and smarter, then I’ve built something that actually matters.

Live Demo: https://hacktracker-production.up.railway.app/
GitHub: https://github.com/srivi19/hacktracker

Built by: Vi (Srividya Narayanan)

Built With

Share this project:

Updates