Sharp11.org is an AI research lab focused on building robust, production-grade trading systems using modern machine learning (Transformers, LLMs, Gradient Boosting, and Reinforcement Learning). Over the past year, we have evaluated recent advances in LLMs, Retrieval-Augmented Generation (RAG), and quantitative ML techniques in live market conditions.

We do not allow ML models to trade autonomously. Instead, we apply a human-in-the-loop framework with a rule-based constraint layer that is reviewed and updated regularly. This approach exists for a reason: market regimes shift, model behavior can drift, and purely automated systems can become brittle under stress.

At Sharp11, AI is a decision-support tool—not an autonomous trader. Human trading knowledge, encoded as explicit risk and execution rules, guides model framing, validation, and deployment to improve robustness, control drawdowns, and reduce operational risk.