Running Autonomous AI Agents with Curupira
How Curupira orchestrates autonomous AI trading agents β from research to execution. Architecture overview, agent lifecycle, monitoring, and how to build your own agent on the platform.
The open-source infrastructure that powers our autonomous trading agents and automated research pipelines. Built with Claude, cron jobs, and 31,000+ lines of strategy code.
Autonomous AI agents that run 24/7, managing their own trading decisions, risk limits, and execution.
Automated backtesting infrastructure. Define a strategy, run it across multiple assets and timeframes, get results.
Real-time market data ingestion from exchanges, on-chain data providers, and alternative data sources.
LLM-powered analysis and decision-making. The agent reads market context and reasons about positions in natural language.
Real-time dashboards for PnL tracking, signal monitoring, and agent health. Alerts for anomalies.
One-click deployment via cron jobs and systemd. Battle-tested on live markets with real capital.
The entire Curupira platform is open source. Fork it, run your own agents, contribute back.
View on GitHub