🚀 Livestream Alert: Build Your Own AI Coach!
Get ready for the creator's first live AI agent build stream this Sunday, October 26th, at 11:00 a.m. Central Time. Unlike previous Q&A sessions, this event will be a hands-on demonstration of building a complete AI agent from scratch, leveraging AI coding assistance. A recording will be available.
The primary project is a personal AI coach, initially trained on the creator's YouTube content using advanced Retrieval-Augmented Generation (RAG) techniques, with plans to expand its knowledge base later.
Key technologies and frameworks for the build include:
- YouTube Data Ingestion: Building a RAG pipeline for extracting metadata and timestamps from YouTube videos, adaptable for any channel.
- AI Agent Framework: Python with Pyantic AI, designed to be LLM-agnostic (Open Router, Olama, Anthropic, OpenAI). It will feature advanced agentic RAG strategies for accurate retrieval and transcript pulling.
- Full Stack Application: Includes token streaming and a complete front-end.
- Database: Supabase for storing YouTube videos, transcripts, embeddings, metadata, and front-end authentication.
- RAG Tool: Dockling for hybrid chunking, critical for accurate YouTube transcript RAG, and future PDF support.
- AI Coding Assistant: Claude Code will be demonstrated, with discussions on compatibility with other AI assistants.
- Task Management & RAG: Integration of the open-source Archon tool to enhance the AI coding workflow.
This livestream also celebrates the release of the "Agentic Coding Course," aiming to provide mastery in AI coding assistance. The goal is a straightforward, practical example of building a functional AI agent using a full AI coding assistant workflow.
Final Takeaway: This stream offers a practical, real-world application of AI agent development using modern tools, providing valuable insights for anyone looking to build robust AI-powered systems with efficiency and precision.
