Tag
作者在 AI Engineer World's Fair 上的 evals 演讲总结，记录了如何将一个 LLM 评审器从完全不可用提升到与人类判断 30 次中有 27 次一致：包括改用点式评分避免偏好偏差、校准上下文、更换模型以及应对谄媚倾向等步骤。
作者探讨独立工程师在AI工具快速发展但行业标准尚未定型的短暂窗口期，如何保持价值与自主性，避免成为平台吸收创意的免费研发力量。
A Philippine software company is sponsoring developers with real AI/LLM project experience to take the official Anthropic Claude Certified Architect exam under a 50/50 cost-share arrangement, including prep materials and mock assessments.
The author strongly recommends an article for anyone building AI applications, automation, and Forward Deployed Engineer roles, outlining the complete AI engineering stack: Python → FastAPI → LLM API → Tool Calling → RAG → Workflow → Agent → Eval → Deployment.
An AI engineering commentary arguing that an AI app is far more than an LLM API call, breaking the stack down into compute/storage, serving, models, context, orchestration, agents, and experience layers — advising builders to start with hosted models and only add complexity as needed.
A practitioner shares a 6-month roadmap to become an AI engineer from zero, covering software foundations, LLM app development, RAG, agents and evals, deployment/reliability, and specialization — emphasizing that companies hire people who build systems on top of models, not people who train them.
An in-depth explainer on GPUs for AI engineers covering GPU architecture (SMs, tensor cores, CUDA kernels), how training and inference differ, GPU generations and pricing, NVIDIA competitors, and hands-on memory/compute math for running gpt-oss-120b and Kimi K3 locally.
Rohit G adds a Projects section to the MIT-licensed AI Engineering from Scratch curriculum, featuring 48 build-from-scratch AI projects across multiple languages with an automated grader that tests every stage.
A tweet lists key AI engineering concepts such as KV cache and speculative decoding, challenging professionals to explain them clearly for interviews in 2026.
Free, open-source AI engineering course with 523 lessons, now available as EPUB/PDF books and in eight languages, focusing on building algorithms from scratch.
This is an open-source textbook published by Harvard University, aiming to establish machine learning systems engineering as a quantitative engineering discipline, similar to computer architecture. The book covers ML system design principles from single nodes to ten-thousand-card clusters and emphasizes engineering rather than just model-centric approaches.
A GitHub repository compiling leaked AI engineering interview questions from 35 major companies, organized by company with direct links to solutions, covering high-frequency topics and interview processes.
An aspiring AI engineer seeks advice from senior professionals on which portfolio project best demonstrates production skills, proposing ideas like LLM routing services, agent systems with guardrails, and local inference microservices, while asking for feedback on common gaps in portfolios.
Stanford has released a free 67-minute course on building AI agents from scratch, covering both coding and no-code methods.
This article elaborates on the key concepts in AI engineering, emphasizing that prompts are the core of engineering aspects such as context, tool integration, loops, evaluation, and graphs, and explains the application of these terms in AI Agent development.
A tweet from @mattpocockuk encourages sharing completion certificates for a course, with a reply from someone seeking to hire course graduates for AI engineering roles at the University of Illinois Chicago.
The team at g2i_ai is organizing the AI Engineer Code Summit in San Francisco this November, bringing together experts in AI-powered software development.
The author shares insights from building an agent harness for GPT-3.5 Turbo, emphasizing that code-based verification and guardrails are crucial for reliable AI agent performance.
This article organizes 13 attention mechanisms in AI by the bottleneck they solve, covering KV cache reduction, attention patterns, compute efficiency, and serving efficiency to help AI engineers understand and apply these techniques.
The tweet celebrates the success of 'The LLM Engineer’s Handbook' as the best-selling book from Packt Publishing in 2025 and teases an upcoming second book.