Ämnen & Kategorier

The DevOps Handbook "The DevOps Handbook," explores the origins and principles of DevOps, advocating for a collaborative approach between Development and Operations teams. It highlights the ineffectiveness of traditional IT silos and processes like manual deployments and end-of-project security reviews, promoting practices such as continuous integration, automated testing, and small batch sizes to improve workflow and reduce risk. The text also discusses the importance of learning from failures through blameless post-mortems and emphasizes the need for visibility across the entire value stream using tools and telemetry. Case studies from companies like Etsy, Nordstrom, and Netflix illustrate the benefits of adopting DevOps, including increased deployment frequency, improved reliability, and better business outcomes.

"Prompt Engineering for LLMs" by John Berryman and Albert Ziegler, provides a detailed guide to building applications with Large Language Models (LLMs). It begins with a history of LLMs and their core architecture, particularly the Transformer and its autoregressive nature, explaining concepts like tokens and logprobs. The text then shifts to practical techniques, discussing how to effectively craft prompts, manage context through methods like retrieval augmented generation (RAG) and summarization, and structure responses. Key aspects of LLM workflows, including the use of tools and the evolution of models through techniques like fine-tuning and Reinforcement Learning from Human Feedback (RLHF), are explored, culminating in strategies for evaluating LLM application quality through both offline and online methods.

A comprehensive deep-dive into the essential engineering practices that power modern AI/ML systems in production. This series takes you from foundational prompt engineering techniques to enterprise-scale MLOps implementations, covering the critical skills every AI engineer, MLOps practitioner, and solution architect needs to build robust, scalable AI systems. Each episode combines cutting-edge research insights with battle-tested industry practices from companies like OpenAI, Google, Netflix, and Uber. Perfect for engineers who want to bridge the gap between AI theory and production reality.