Zerve, an AI-native data science platform, is hosting the ZerveHack hackathon. The central challenge invites participants to leverage the platform's agent-driven capabilities to move from a question to a complete analysis. Competitors are tasked with building an original project by analyzing a dataset to uncover insights and deploying it as a shareable analysis, application, or API. The goal is to demonstrate what becomes possible when AI handles the coding and execution, allowing the user to focus on steering the analytical direction.
Program Timeline
• Submission Period: February 26, 2026 (9:00 am Eastern Time) – April 29, 2026 (2:00 pm Eastern Time)
• Judging Period: April 30, 2026 (2:00 pm Eastern Time) – May 8, 2026 (2:00 pm Eastern Time)
• Winners Announced: On or around May 13, 2026 (10:00 am Eastern Time)
Prizes
• Total Prize Pool: $10,000 in cash prizes are available.
• 1st Place: $5,000 in cash for one winner.
• 2nd Place: $3,000 in cash for one winner.
• 3rd Place: $2,000 in cash for one winner.
Challenge Overview
• Core Task: Use the Zerve platform to build an original project. This involves analyzing a dataset, uncovering insights, and shipping a usable product such as an analysis, an app, or an API.
• Methodology: There are no constraints on the topic or prescribed methodology. Participants are encouraged to bring a question and explore it using the platform.
• Focus: The best submissions are expected to showcase the platform's strengths, demonstrating what is possible when AI handles execution while the user drives the direction, moving beyond traditional manual queries and disconnected tools.
Project Inspiration and Datasets
• Prediction Markets & Forecasting: Analyze what market odds indicate and where they might be wrong. Suggested data sources include the Polymarket API, Kalshi API, and Metaculus. Potential projects could involve calibration analysis, arbitrage detection APIs, or live odds tracking apps.
• Finance & Economics: Find signals within economic data. Suggested data sources include FRED, Alpha Vantage, and SEC EDGAR. Potential projects could involve leading indicator discovery, sector rotation analysis, or macro regime detection.
• Sports & Competition: Settle debates using rich historical sports data. Suggested data sources include FBref, the NBA API, and Stathead. Potential projects could involve World Cup 2026 predictions, player valuation models, or in-game win probability APIs.
• Climate & Energy: Address high-stakes questions related to weather, energy, and emissions. Suggested data sources include NOAA Climate Data, EIA, and EPA Air Quality. Potential projects could involve extreme weather prediction, renewable energy forecasting, or grid reliability analysis.
• Social & Cultural Trends: Quantify what people are searching for and talking about. Suggested data sources include Google Trends, Hugging Face Datasets, and Reddit Data. Potential projects could involve trend prediction, sentiment analysis, or demographic shift analysis.
• Wildcard (Bring Your Own Data): Participants are encouraged to use their own unique datasets from work, hobbies, or personal interests.
Submission Requirements
• Public Zerve Project: A complete, published, and shareable analysis that runs without errors.
• Project Summary: A description of no more than 300 words answering: What question did you ask? What did you find? Why does it matter?
• Demo Video: A video of no more than three minutes walking through the project, showing the workflow, highlighting findings, and explaining the approach. The video must be publicly visible on a platform like YouTube or Vimeo.
• Social Media Post: A post on a platform like LinkedIn or X that includes a link to the project, a message about competing, a brief description of the project, and tags for @Zerve AI (LinkedIn) or @Zerve_AI (X).
• Deployment (Highly Encouraged): While not required, deploying the project as a live API or an interactive application is highly encouraged and will receive priority consideration from judges.
• Project Originality: Projects must be newly created or significantly updated during the hackathon's submission period.
Judging Criteria
• Analytical Depth (35%): Assesses if the project goes beyond surface-level exploration to provide genuine insight that changes how one thinks about the data.
• End-to-End Workflow (30%): Evaluates the use of Zerve's autonomous reasoning to move from question to answer, with higher scores for projects taken to production as deployed APIs or apps.
• Storytelling & Clarity (20%): Judges whether the project's findings and importance can be understood by an external audience.
• Creativity & Ambition (15%): Considers if the project took risks, combined unexpected data sources, or asked a novel question.
Eligibility
• Eligible Participants: The hackathon is open to individuals aged 18 or older who are legal residents of the 50 United States (including D.C.) or Canada (excluding Quebec), as well as teams of eligible individuals and organizations incorporated in the U.S. or Canada.
• Ineligible Participants: The hackathon is not open to residents of other countries, organizations involved in the production or promotion of the event, their employees and immediate family, judges, or any individual or organization that would create a conflict of interest.
The ZerveHack hackathon presents a unique opportunity for data scientists, developers, and analysts to explore the frontiers of AI-native data science. By providing the tools to automate complex coding and analysis, the event challenges participants to focus on creative problem-solving and delivering impactful, production-ready insights and applications. Interested parties are encouraged to review the full eligibility and submission details to participate.