Automated Investing

Overview

Explore using always-on AI agents for automated investment strategies. The thesis: trading is a data-intensive domain with clear feedback mechanisms — exactly the kind of problem an agent that runs 24/7 can solve better than a human who can’t watch markets constantly.

Why This Fits Ray Data Co

• Data pipeline → signal detection → decision → execution → feedback → compound learning
• Same agentic infrastructure we’re building for everything else
• Clear, measurable feedback loop (P&L is the most honest signal)
• Always-on agent reacts to signals at the right time — something the founder can’t do manually

What We’re NOT Doing

• Not endorsing any specific trading strategy yet
• Not starting with real money before we have a tested system
• Not building a day-trading bot — exploring systematic, data-driven approaches

What We Need to Learn

• Regulatory requirements (SEC, broker APIs, pattern day trader rules)
• Broker APIs for automated execution (Alpaca, Interactive Brokers, TD Ameritrade)
• Backtesting frameworks and historical data sources
• Risk management and position sizing
• What strategies are actually viable for algorithmic/agentic trading at small scale

Reference Material

• 06-reference/2026-04-04-swing-trading-guide — Kevin Xu’s swing trading approach (volume + catalysts)
• 06-reference/2026-04-10-gemchange-quant-from-scratch — @gemchange_ltd’s 18-month curriculum to go from zero to quant (probability → stats → linalg → calc → stochastic calc). The math foundation this project needs to sit on top of.
• 06-reference/2026-04-10-gemchange-simulate-like-quant-desk — @gemchange_ltd’s implementation guide for prediction-market simulation (Monte Carlo → importance sampling → particle filters → copulas → agent-based models → 5-layer production stack). Direct blueprint for the engine we build on top of the roadmap.

Discord Channel

TBD — add when ready to activate