Kristian Fagerlie describes using AI coding and analysis toolsCode generation uses AI or another automated system to create source code from instructions, examples, schemas, or higher-level specifications. to explore trading ideas as separate experimental workflows, which he calls pods. He discusses sourcing market data, evaluating ideas and monitoring the systems.
Fagerlie demonstrates an analysis of historical relationships between pairs of stocks and asks a model to explain potential signals. He also describes a short observation of a prediction-market experiment and scheduled monitoringWorkflow automation uses software or AI to complete a connected sequence of routine steps with less manual effort..
The examples are his own experiments and do not establish a durable trading edge. He stresses understanding the model’s output before deploying a strategyHuman-in-the-loop AI keeps a person involved in reviewing, correcting, approving, or guiding an AI system's work.; historical relationships and short test periods do not guarantee future performance.
Watch on YouTube
