A full-stack platform for discovering and exploring BJJ techniques. Built with Next.js, MongoDB, and machine learning—powered trend analysis to surface the most relevant grappling content.
Custom ETL pipeline ingests and normalizes video metadata, timestamps, and hashtag taxonomies from 23,000+ BJJ instructional clips stored in MongoDB.
A trending score algorithm analyzes technique frequency, recency, and cross-reference density to surface emerging grappling concepts in real time.
MongoDB Atlas text indexes with autocomplete suggestions enable instant search across notes, hashtags, and technique descriptions.
Browse, search, and filter across the full video database