×
It does now. And we'll pay $10,000 to the people who prove it.
Five data domains with proven participation. Use public datasets, your own data, or both. You just have to build something real that runs on Zerve.
Polymarket, Kalshi, Metaculus. Build models that find signal in probability markets, news sentiment, and event data.
FRED, Alpha Vantage, SEC EDGAR. Backtesting, macro signals, sector rotation. Build the analysis quant researchers actually run.
Google Trends, Hugging Face, Reddit. Find the signal before it becomes obvious. Trend prediction, sentiment, cultural moment detection.
The best submissions are usually personal obsessions. Your dataset, your question, your output. The only requirement is that it runs and deploys on Zerve.
Cursor, Copilot, Claude Code — they help you write faster. Zerve does something different. It orchestrates the full loop that data work actually requires.
People building on Zerve work at
You write the code. Then you run it somewhere else. Then you copy the output back. Then you write the next step. You have been doing this for years.
Describe what you want. Zerve writes the code, runs the compute, captures what the last block produced, uses that to decide what comes next, and keeps going until the analysis is done. You review. You steer. You stop assembling.
“I gave it a churn analysis task and it pulled data, handled missing values, ran segmentation, and scheduled the workflow automatically — using context from my past work.”
Eric Vyacheslav — AI EngineerYou want to test 100 parameter combinations. So you test them one at a time. You have been doing this since your first job.
Fleet distributes your sweep across parallel compute automatically. You describe the parameter grid. Zerve fans it out. You stop thinking in individual runs and start thinking in sweeps.
“That shift alone saved us a huge amount of time and let us move faster on model development.”
Arindam Bose — Head of Data Science, Sun KingYou finish the analysis. You hand it to engineering. You wait. Three months later, something ships. Maybe.
Zerve produces a Streamlit app, an API endpoint, or a conversational report that stakeholders can query — as a natural output of the analysis you just ran. You do not change tools. You do not wait for a sprint. You just hit deploy.
“Zerve feels different because it was built for data scientists from the ground up. It understands context, adapts as you work, and lets you iterate without crashes or lost work.”
Eduardo Ordax — Gen AI Lead at AWSSomeone leaves. Their notebooks are unrunnable. Six months of work, gone. You have rebuilt from scratch before and you will again.
Zerve holds the context of your analysis — schema, methodology, variables, prior results — across every block and every session. A teammate can pick up where you left off. A new hire can build on it without a handover meeting.
“The Zerve agent has transformed our data science workflow. Backlogs are vanishing and we're experimenting faster than ever.”
Mike Smith — CEO, Ascendr.orgThree top prizes. Cash rewards for the top 100. You don't need a breakthrough — you just need to build something real and run it on the platform.
1st place
Grand Prize
The most rigorous, well-documented project that deploys a real output — an API, an app, or a report others can use. Any track. Any complexity level that earns it.
2nd place
Runner Up
Strong methodology, clear documentation, and a deployed or deployable output. The kind of project you would be proud to show a colleague.
3rd place
Third Prize
A clear question, a clean workflow, results that someone else could build on. Does not need to be complex. Needs to be honest work.
You don't have to finish on the podium to get something out of this. Every submission in the top 100 that shows genuine platform engagement — running workflows, using credits, producing an output — gets a reward package. Simple projects are eligible.
Watch the platform in action, read the docs, ask for help in Slack.
Sign up, build, ship, share. Reach out if you need help or want to give feedback.
Sign up free. 300 credits instantly. No promo code needed. No card.
Open a sample project from the gallery, or start from a blank Zerve workspace.
Build & deploy. Swap in your data, run agents, publish your Zerve project.
Share & get support. Post about your project and reach out for help or feedback.
Submissions close Sept 13 · 11:59 PM
READY WHEN YOU ARE
Sign up free. Get 300 credits instantly. No promo code needed. Describe what you want to build. See what happens.
×300 free credits · No card needed · $10,000 in prizes
