How to create your own AI performance coach: Optimizing your unique nutrition, recovery, and injury management needs | Lucas Werthein (Cactus)

What It Covers

Lucas Werthein demonstrates how he built a custom ChatGPT performance coach using MRIs, blood tests, WHOOP data, nutrition plans, and injury records to optimize athletic performance, recovery, and injury management at age 40.

Key Questions Answered

• •Multi-format data integration: ChatGPT processes MRIs, X-rays, blood tests in PDF format, CSV files from wearables, nutrition plans, and journal entries across multiple languages without requiring data cleaning or standardization, enabling comprehensive health synthesis.
• •Prompt architecture for coaching: Effective GPT prompts combine role definition, realistic outcome goals, hard boundaries on what not to recommend, tone specifications, and cross-validation requirements rather than demanding extreme optimization or unproven biohacking interventions.
• •Visual input for diagnosis: Taking photos or videos of pain points, circling specific areas, and uploading to ChatGPT enables on-demand validation of expert opinions and personalized explanations in accessible formats when medical professionals are unavailable.
• •Synthetic expert workflows: Creating GPTs trained on publicly available information about clients or colleagues enables teams to get 80-90% accurate responses to questions when those experts are unavailable, reducing meeting overhead in distributed work environments.