What is Epsilab?

Epsilab is an AI-powered quantitative research platform that transforms natural language trading ideas into systematic strategies, runs comprehensive backtests with realistic market conditions, and automatically optimizes parameters—all through a simple interface.

Who is it for?

Epsilab is designed for quantitative researchers, algorithmic traders, and portfolio managers who want to:

• Generate strategies from natural language descriptions
• Automatically optimize parameters through systematic testing
• Run systematic backtests with realistic transaction costs and market impact
• Aggregate multiple strategies into risk-managed portfolios
• Conduct iterative research through AI-powered strategy improvement

How It Works

From trading idea to live portfolio:

1. Describe → Enter your trading idea in natural language
2. Generate → AI creates and backtests your strategy automatically
3. Optimize → System finds the best parameters for your strategy
4. Research → Iteratively improve through multiple variations
5. Deploy → Add winning strategies to your live portfolio

Platform Architecture

Epsilab uses a distributed architecture designed for security, privacy, and scalability:

Trade Servers (Per-User)

Each user receives their own dedicated trade server—an isolated compute instance that handles strategy research, backtesting, and live portfolio execution. Your server scales automatically based on workload and idles when not in use. All strategy code and data remain isolated in your dedicated environment.

Pod Aggregators (Shared Infrastructure)

Pod aggregators handle multi-user portfolio construction while preserving privacy. They only see strategy outputs (target positions), never source code—enabling collaborative investing while protecting your intellectual property. Aggregators combine signals, apply risk constraints, and coordinate live execution across strategies.

What's Supported

Epsilab currently supports:

• Markets: US equities (stocks and ETFs) and currencies (crypto + forex)
• Data: Daily market data and corporate actions
• Strategy Types: Technical, fundamental, factor-based, and quantitative approaches
• Timeframes: Multiple rebalancing frequencies from intraday to monthly

Getting Started

Ready to build your first strategy? Head over to our Quick Start guide to walk through the complete process from idea to deployment.