Check back later.
You must be a hackathon organizer or judge to access that function.
Theme
Tame the chaos of autonomous finance by building agents that can trade, reason, and integrate market data in real time. Leverage APIs like Alpha Vantage and DeFi protocols, and fuse predictive ML models with robust portfolio logic. Deliver a self‑driving money engine that reshapes how capital moves across assets, all in a single hack day.
Detailed Schedule
Judging Criteria
These are the judging criteria we’ll use to evaluate projects.
| Criterion | What’s Evaluated | Score (1‑5) |
|---|---|---|
| Running Code | Does the submission deliver a live, deployable agent that actually trades, fetches market data, or executes finance logic? | 1 – No runnable code, only slides or mockups. 2 – Code runs but crashes or fails core finance functions. 3 – Core flow works but major bugs or missing key features. 4 – Mostly stable; minor UI glitches or occasional API hiccups. 5 – Fully functional, zero‑error demo, can run unattended. |
| Innovation & Creativity | How novel is the approach to autonomous finance? Are new algorithms, cross‑asset reasoning, or unique integrations used? | 1 – Straight‑forward copy of an existing tutorial. 2 – Minor twist on a well‑known pattern. 3 – Some original component but largely derivative. 4 – Clear novel idea (e.g., hybrid stock‑crypto risk model) with creative implementation. 5 – Groundbreaking concept that re‑imagines how agents manage money, with original architecture. |
| Real‑world Impact | Potential for the agent to deliver measurable financial value or solve a tangible market problem. | 1 – No practical use; purely academic. 2 – Concept could be useful but lacks clear path to impact. 3 – Solves a niche need; moderate ROI potential. 4 – Addresses a significant pain point (e.g., automated portfolio rebalancing) with demonstrable ROI. 5 – High‑stakes impact (e.g., live arbitrage bot that consistently outperforms benchmarks). |
| Theme Alignment | How well does the project embody the “autonomous finance agents” theme, including multi‑asset reasoning, API integration, and self‑driving money flows? | 1 – Barely related; generic AI app. 2 – Mentions finance but no autonomous agent behavior. 3 – Uses a finance API but lacks self‑driving logic. 4 – Implements an autonomous agent that trades or reasons across assets using at least one API. 5 – Fully realizes the theme: an end‑to‑end autonomous money engine that fetches data, makes decisions, executes trades, and learns from outcomes. |
Handbook View Handbook
Hackathon Participant Handbook
Introduction
Welcome to our AI hackathon! This event is all about building innovative AI-powered applications with the help of cutting-edge development tools. Your mission is to create an AI-driven app that’s original, useful, or even delightfully weird – it could be a voice assistant, a finance agent, a crypto trading bot, or anything you can imagine. We want to see you push AI beyond the ordinary and ship something amazing. This handbook will guide you through the event details and provide a ton of free resources (APIs, datasets, and more) to power your project. The key requirement is that all teams use LiquidMetal AI’s Raindrop platform and Claude Code for development, ensuring you get hands-on experience with these state-of-the-art tools.
Event Overview
Hackathon Format
This hackathon brings together developers, builders, and tinkerers to rapidly prototype AI-native applications. You can participate solo or in a team of up to 4 people. Don’t worry if you don’t have a team – we’ll help match you with others, so everyone can collaborate and contribute.
Tools & Platform
Each project must be built on the Raindrop Platform and developed using an AI coding assistant (Claude Code). Raindrop is the first framework designed for coding agents rather than humans – it lets you build and deploy production-ready apps through AI assistance. Claude Code, Anthropic’s powerful coding AI, works directly with Raindrop’s MCP (Model Context Protocol) server to help you architect, code, and launch your application faster. In practice, this means you’ll be writing natural-language instructions to Claude, which will generate and manage code/infrastructure on Raindrop for you.
Getting Started
Raindrop’s platform provides Smart Components – SmartBuckets, SmartSQL, SmartMemory, SmartInference – that you can leverage in your app. By describing your application and requirements to Claude Code, you can generate everything from database schemas to API endpoints, with Raindrop handling deployment and scaling automatically.
Hackathon Schedule
The event kicks off with an opening ceremony and a quick workshop on using Raindrop and Claude Code. Then it’s heads-down building time until the final submission deadline, when you’ll demo your working app to judges.
Submission Requirements
To qualify for judging, you’ll need to submit:
- A link to your live deployed app (Raindrop provides URLs for projects)
- A short demo video (2–3 minutes)
- Access to your source code (GitHub repo or ZIP)
Judges will evaluate based on innovation, real-world usefulness, technical execution, and creative use of AI.
Technical Requirements & Guidelines
Use of Raindrop
Your application’s backend must run on LiquidMetal AI’s Raindrop platform. Raindrop is “Claude-native” – it’s designed to be operated by AI agents (like Claude) via the Model Context Protocol. Raindrop sets up databases, APIs, authentication, and more, guided by Claude Code.
Using Claude Code
Participants must use an AI coding assistant, ideally Claude Code. Treat Claude as a junior developer: be clear, test results, and iterate. You maintain control over what gets deployed.
Additional Tools
You can use any frameworks or libraries in Raindrop (e.g. Python libraries for finance or APIs). For frontend hosting, use Netlify or
Vercel. For authentication, try
WorkOS (free tier for 1M users). For payments,
Stripe test mode is available.
Free APIs and Data Resources
Free Financial Market APIs
-
Alpha Vantage — Free stock, forex, and crypto data with built-in indicators.
-
Yahoo Finance (yfinance) — Python API for stock data.
-
Finnhub — Real-time quotes, company news, and sentiment.
-
IEX Cloud — Free real-time U.S. market data.
-
Twelve Data — Global equities, crypto, and indicators.
-
Polygon.io — Stocks, options, forex, and crypto feeds.
Cryptocurrency APIs
-
CoinGecko — No API key required, extensive market data.
-
CoinMarketCap — Basic tier for quotes, market caps.
-
CoinPaprika — Free crypto API, no key required.
-
DexScreener — Real-time DEX pair info across 80+ blockchains.
-
QuickNode,
Alchemy — Blockchain RPC APIs.
Open Financial Datasets
-
data.gov — U.S. datasets.
-
data.gov.in — Indian government datasets.
-
data.europa.eu — European open data.
World Bank Open Data
FRED
SEC Edgar
IMF Data
Nasdaq Data Link
Kaggle Datasets
Papers With Code – Finance
Open-Source Tools for Finance
-
CCXT — Crypto exchange trading library.
-
Freqtrade — Open-source crypto trading bot.
-
Hummingbot — Market-making bot framework.
-
Backtrader — Backtesting trading strategies.
-
Alpaca — Paper trading API for stocks.
-
-
TA-Lib — Technical analysis library.
Other Free Services
- Hosting:
Netlify,
Vercel
- Auth:
WorkOS,
Firebase Auth
- Payments:
Stripe
- Cloud Credits:
AWS Free Tier
Closing Notes
Use these resources to create something truly original. The best projects combine strong engineering with creativity and real-world impact. Good luck and happy hacking!