Training Science Podcast AI Agents, HRV Readiness & the Next Era of Training: A Deep Dive with Dr. Andrea Zignoli & Prof Paul Laursen
Dec 5, 2025
Dr. Andrea Zignoli, AI modeling lead at Athletica and sport science researcher, dives into the revolutionary role of AI in endurance training. He discusses his recent publications on AI agents and HRV readiness monitoring, exploring how AI can enhance athlete decision-making by interpreting training data. The conversation also covers the concept of 'Sport Science 3.0,' where AI supplements human coaching, and the significance of sentiment analysis as an internal load signal. Zignoli emphasizes the importance of ethically integrating AI into sports science.
AI Snips
Chapters
Books
Transcript
Episode notes
Integration Is The Next Phase
- Sports Science 3.0 bridges coach intuition and automated data processing using AI.
- Andrea says this fusion will drive the next decade of research and practice.
Knowledge As Trajectories
- LLMs represent text as high-dimensional vectors forming knowledge trajectories.
- Andrea visualized and compared human and model knowledge pathways to show similar topical breadth.
Ground LLMs With Domain Priors
- Combine domain priors with RAG to ground LLM outputs and reduce hallucinations.
- Use established HRV baselines and let the model explain results, not replace core algorithms.
