Building With LLMs

As large language models (LLMs) transition from research projects to enterprise workflows, this chapter aims to be a practical guide to implementing them through the solution lifecycle, from exploration through deployment and monitoring. There are considerations regarding architecture decisions, such as whether to use an API or self-hosted model, level of integration in the system, and trade-offs between costs, performance, and scalability. Also examined are application approaches, such as fine-tuning, prompt engineering, and RAG, to name a few, in addition to cloud and open-source platforms, MLOps practices, and real-world examples and case studies, to ultimately support technical and business leaders in capitalizing on the potential for LLMs to create sustained impact.