LLMs in Production

Engineering AI Applications

By: Christopher Brousseau, Matt Sharp

Narrated by: Christopher Kendrick

Unlock the potential of Generative AI with this Large Language Model production-ready playbook for seamless deployment, optimization, and scaling. This hands-on guide takes you beyond theory, offering expert strategies for integrating LLMs into real-world applications using retrieval-augmented generation (RAG), vector databases, PEFT, LoRA, and scalable inference architectures. Whether you're an ML engineer, data scientist, or MLOps practitioner, you’ll gain the technical know-how to operationalize LLMs efficiently, reduce compute costs, and ensure rock-solid reliability in production.

What You’ll Learn:

• Master LLM Fundamentals – Understand tokenization, transformer architectures, and the evolution linguistics to the creation of foundation models.
• RAG & Vector Databases – Augment model capabilities with real-time retrieval and memory-optimized embeddings.
• Training vs Fine-tuning – Learn how to train your own model as well as cutting edge techniques like Distillation, RLHF, PEFT, LoRA, and QLoRA for cost-effective adaptation.
• Prompt Engineering – Discover the quickly evolving world of prompt engineering and go beyond simple prompt and pray methods and learn how to implement structured outputs, complex workflows, and LLM agents.
• Scaling & Cost Optimization – Deploy LLMs into your favorite cloud of choice, on commodity hardware, Kubernetes clusters, and edge devices.
• Securing AI Workflows – Implement guardrails for hallucination mitigation, adversarial testing, and compliance monitoring.
• MLOps for LLMs – Learn all about LLMOps, automate model lifecycle management, retraining pipelines, and continuous evaluation.

Hands-on Projects Include:

• Training a custom LLM from scratch – Build and optimize an industry-specific model.

• AI-Powered VSCode Extension – Use LLMs to enhance developer productivity with intelligent code completion.

• Deploying on Edge Devices – Run a lightweight LLM on a Raspberry Pi or Jetson Nano for real-world AI applications.

©2024 Manning Publications (P)2025 Manning Publications