They'll Fly You to Vegas if You Win This Coding Challenge

This tutorial explains the architecture and implementation of an application designed to predict the outcomes of robot combat matches. The creator demonstrates how to gather data from BattleBots websites and Reddit using Bright Data’s Web Unlocker API to bypass scraping protections. The pipeline involves parsing this data, storing it in a local SQLite database, and vectorizing the information to enable Retrieval-Augmented Generation (RAG). By passing this retrieved context and specific stats into an LLM, the system generates a structured scouting report with a predicted winner. The video serves both as an example of a full-stack project for viewers to study and as a promotional overview of a BattleBots developer competition that challenges builders to create a similar project using provided data.