Building Blocks for LLM Systems and Products - Seven Key Patterns for Integration
AI Engineer via YouTube
Free AI-powered learning to build in-demand skills
AI Adoption - Drive Business Value and Organizational Impact
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn practical patterns for integrating large language models into production systems and products through this 17-minute conference talk. Discover seven key building blocks organized along the spectrum of improving performance versus reducing cost and risk, spanning from data-centric to user-facing approaches. Explore evaluations for measuring performance, retrieval-augmented generation (RAG) for incorporating recent external knowledge, and fine-tuning techniques for task-specific improvements. Master cost and latency optimization through caching strategies, implement guardrails to ensure output quality, and design defensive user experiences that gracefully handle errors. Understand how to establish data flywheels through systematic user feedback collection, drawing insights from academic research, industry resources, and real-world practitioner experience to bridge the gap between impressive demos and robust production systems.
Syllabus
Introduction
Evaluations
Guard Rails
Collecting Feedback
Summary
Taught by
AI Engineer