What is Zerve?

Zerve is an AI-native data science platform. Start with a question, not a blank script. Zerve's agent writes the code, iterates on errors and runs the analysis while you steer. When you're ready, deploy it as an API or app without leaving the platform.

The Challenge

Use Zerve to build something original. Analyze a dataset, uncover insights, and ship something others can use: an analysis, an app, an API, or all three.

No constraints on topic. No prescribed methodology. Bring a question and see how far you can take it.

The best submissions will feel like they couldn't have been built the old way. Manual queries, disconnected tools, endless waiting on the data team. Show us what's possible when AI handles the execution and you drive the direction.

Getting Started 

Navigate to the resources tab to signup and get started! 

Requirements

What You Could Build

We've curated datasets across five areas. Pick one, combine a few, or ignore them entirely and bring your own.

Prediction Markets & Forecasting

Markets are betting on everything: elections, Fed decisions, AI breakthroughs, celebrity gossip. What do the odds actually tell us? Where are they wrong?

  • Polymarket API | Crypto-native prediction market with broad coverage
  • Kalshi API | Regulated US prediction market
  • Metaculus | Forecasting community with historical accuracy data

Possible angles: Calibration analysis (how accurate are markets?), arbitrage detection APIs, sentiment correlation with news events, live odds tracking apps

Finance & Economics

Tariffs, rate cuts, inflation, employment. Plenty of noise, plenty of signal if you know where to look.

  • FRED | Federal Reserve economic data (800k+ time series)
  • Alpha Vantage | Stock, forex, crypto data
  • SEC EDGAR | Public company filings

Possible angles: Leading indicator discovery, sector rotation analysis, earnings sentiment vs. stock performance, macro regime detection

Sports & Competition

World Cup 2026 is this summer. Plus NBA playoffs, MLB season, golf majors. Rich historical data, clear outcomes, endless debates to settle.

  • FBref | Comprehensive soccer/football statistics
  • NBA API | Player and game data
  • Stathead | Multi-sport historical data

Possible angles: World Cup 2026 predictions (build a live prediction app), player valuation models, in-game win probability APIs, draft pick analysis

Climate & Energy

Weather patterns, grid stress, emissions tracking. High-stakes, data-rich, and increasingly relevant.

Possible angles: Extreme weather prediction, renewable energy forecasting, grid reliability analysis, emissions attribution

Social & Cultural Trends

What people search for, talk about, and care about—quantified.

Possible angles: Trend prediction, sentiment analysis, demographic shifts, cultural moment detection

Wildcard: Bring Your Own Data

The best projects often come from personal obsessions. Got a dataset from work (anonymized), a hobby, or a question that's been nagging you? Use it.

 

What to Submit

Required:

  1. Public Zerve Project | Your complete analysis, published and shareable. Must run without errors.
  2. Project Summary (300 words max) | What question did you ask? What did you find? Why does it matter?
  3. Demo Video (3 minutes max) | Walk us through your project. Show the workflow, highlight your findings, explain your approach.
  4. Share it | Post your project on LinkedIn, X or your platform of choice. Tag @Zerve AI (LinkedIn) or @Zerve_AI (X). We'll amplify the best ones.

Bonus (Highly Encouraged):

Deploy It | Don't just analyze. Ship something others can use.

  • API Deployment: Turn your model or analysis into a live endpoint
  • App Deployment: Build an interactive application your audience can explore

Deployed projects will receive priority consideration from judges. Analysis is good. Production is better.

 

Hackathon Sponsors

Prizes

$10,000 in prizes
1st Place
$5,000 in cash
1 winner

2nd Place
$3,000 in cash
1 winner

3rd Place
$2,000 in cash
1 winner

Devpost Achievements

Submitting to this hackathon could earn you:

Judges

Zach O'Hagan

Zach O'Hagan
Growth & GTM | Zerve AI

Greg Michaelson

Greg Michaelson
Co-founder & CPO | Zerve AI

Judging Criteria

  • Analytical Depth (35%)
    Did you go beyond surface-level exploration? We're looking for genuine insight. Something that changes how you think about the data, not just describes it.
  • End-to-End Workflow (30%)
    Did you use Zerve's autonomous reasoning to move efficiently from question to answer? Did you take it all the way to production? Deployed APIs and apps demonstrate the full power of the platform, and score higher here.
  • Storytelling & Clarity (20%)
    Can someone who didn't build this understand what you found and why it matters? Your project should speak for itself.
  • Creativity & Ambition (15%)
    Did you take a risk? Combine unexpected data sources? Ask a question nobody else thought to ask?

Questions? Email the hackathon manager

Invite others to compete

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.