GitHub - eugeneyan/applied-ml

Abstraction: Curated industry ML blog posts and papers from production deployments

Key points:

• Covers 30+ categories: data quality, feature stores, classification, recommendation, search/ranking, NLP, CV, RL, anomaly detection, graph, MLOps, ethics, and team practices
• Feature store section covers Michelangelo Palette (Uber), Feast (Gojek/Google), Feathr (LinkedIn), and DoorDash Riviera
• Recommendation systems span from Amazon item-to-item collaborative filtering (2003) through deep learning approaches at Netflix, YouTube, Spotify, TikTok/ByteDance, and Alibaba
• Companies represented include Airbnb, Netflix, Uber, LinkedIn, Google, Amazon, Meta, Spotify, Twitter, DoorDash, Pinterest, and many others
• MLOps topics include model management, A/B testing, efficiency, infrastructure, and organizational practices
• Maintained by Eugene Yan; companion repos include ml-surveys and applyingML

Connections: Github · Netflix · Uber · Machine Learning · Mlops · Recommender Systems · Feature Stores

Source: https://github.com/eugeneyan/applied-ml