Marc Wyatt

Successfully deploying LLMs on the trading desk hinges on robust data infrastructure—not just model performance. This panel dives into the data engineering required to support LLMs in production, from clean input capture to curated datasets built for the speed and specificity of trading environments. 

• How to capture and process high-frequency trading data for LLM applications
• Best practices for data labeling, storage design, and real-time pipeline optimization
• Meeting governance, auditability, and compliance standards when working with sensitive data; and
• Practical strategies for making LLMs operationally viable in fast-moving, regulated trading settings