Explores a futuristic e-commerce concept where AI translates user prompts into custom-built products, coordinating a network of specialized vendors.
Part two of building a personal recommendation system, covering data collection from Pocket and content extraction using the Jina Reader API.
A developer documents the first steps in building a personalized content recommendation system using saved articles, text embeddings, and algorithms.
Explores how large language models (LLMs) are transforming industrial recommendation systems and search, covering hybrid architectures, data generation, and unified frameworks.
A summary of a talk on applying Large Language Models (LLMs) to build and deploy recommendation systems at scale, presented at Netflix's PRS workshop.
Explores user interfaces for LLMs that minimize text chat, using clicks and user context for more intuitive interactions.
Explores counterfactual evaluation as an alternative to A/B testing for offline assessment of recommendation systems.
Explores how reinforcement learning methods like bandits and policy-based approaches can improve recommendation systems by optimizing for long-term rewards.
Explores system design patterns for industrial-scale recommendation and search engines, focusing on offline/online components and retrieval/ranking stages.
Explores machine learning patterns like bandits, sequential, and graph-based models for personalizing recommendations and search results.
A technical deep dive into real-time machine learning for recommendation systems, comparing approaches in China and the US and discussing implementation.
An interview with an Amazon Applied Scientist describing the daily work, challenges, and projects involved in building ML systems like book recommendations.
Explains four levels of customer targeting, from no segmentation to advanced recommendation systems, and their business applications.
Analyzes the flaws in Hacker News and Reddit ranking algorithms and proposes a randomized solution to improve content discovery.
A developer explores the idea of using social graphs to find interests and people completely unrelated to their own profile.
