Analysis of Agentic RAG Performance vs. Gemini for the 2025 CrossFit Games Rulebook

Join Nick, co-founder and CEO of Trieve, as he demonstrates the power of Agentic RAG (Retrieval-Augmented Generation) through a hands-on tutorial and live testing.

In this video, Nick walks through:

🔍 What is Agentic RAG?

- How it allows large language models to write their own search queries
- The iterative process of refining search terms for better results
Why it outperforms traditional RAG approaches

💻 Practical Implementation

- Step-by-step guide from their new blog post "How to Build a Gentech Rag - or Any PDF in 10 Minutes"
- Ready-to-use CLI tool with complete source code
- Compatible with Cursor, Claude, and other AI coding assistants

🥇 Real-World Testing

Using the 2025 CrossFit Games rulebook as a benchmark dataset, Nick demonstrates how Agentic RAG successfully answers complex queries that stumped traditional RAG systems:

- Prohibited camera lenses
- Team composition rules
- Workout penalties and consequences
- Competition regulations and procedures

📊 Results

The demo shows Agentic RAG achieving 100% accuracy on queries where traditional RAG failed, while providing more detailed, referenced answers than simple context window approaches.

Perfect for developers interested in advanced RAG implementations, AI search optimization, and practical large language model applications. All code and implementation details are available in the accompanying blog post.

Видео Analysis of Agentic RAG Performance vs. Gemini for the 2025 CrossFit Games Rulebook канала Trieve