No. 434 of 500 No. 54 in Generative AI & LLMs

LLMs in Production

From language models to successful products

By Christopher Brousseau and Matthew Sharp

Print length
456 pages
Publisher
Manning
Reading level
Advanced
ISBN-13
978-1633437203
Published
February 11, 2025
Written for
ML engineers, data scientists and platform teams responsible for deploying and operating LLM applications.

Synopsis

Christopher Brousseau and Matthew Sharp, engineers who have led large-scale language model deployments, wrote this Manning book for teams that need LLMs to run reliably outside the lab. The early chapters explain how language models work, from linguistic foundations to transformer architecture, and then address the decision every team faces: whether to call a hosted model, adapt an open one or train its own. From there the authors cover LLMOps platforms, data engineering for training sets, training and fine-tuning methods such as LoRA, and the practicalities of serving very large models at acceptable cost.

Later chapters treat prompt engineering, application design, agents and tool use, and the direction of the field, including knowledge editing and regulation. Three projects tie the material together: reimplementing Llama 3 and training a small model, building a coding assistant extension for VS Code, and deploying a compact model to a Raspberry Pi. Readers come away with a map of the full LLM lifecycle, a shared vocabulary for discussing cost, latency and evaluation with colleagues, and worked examples they can adapt for their own deployments.

Scorecard

Editorial score

How it scored

Editorial score: 65.3 / 1005 weighted criteria

  • Influence30%

    16.5/30
  • Authority25%

    16.3/25
  • Practical value20%

    15.0/20
  • Currency15%

    10.5/15
  • Readability10%

    7.0/10
  • Editorial score

    65.3/100

How the rank was computed

Rank No. 434 of 500 comes from a composite of 48.9: observable signals 19.5 of 55 points plus editorial review 29.4 of 45 points.

All 11 inputs
InputPoints
Observable signals19.5/55
Reach (Amazon ratings volume)5.8/16.5
Consensus across best-of lists5.5/13.8
Awards and bestseller evidence0.0/8.3
Reader rating, sample-adjusted5.7/8.3
Amazon sales rank0.8/5.5
Recency1.7/2.8
Editorial review29.4/45
Author authority7.3/11.3
Influence7.4/13.5
Practical value6.8/9.0
Readability3.2/4.5
Currency4.7/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

A broad engineering guide to taking large language models from prototype to production, strongest on infrastructure, serving and fine-tuning trade-offs.

Strengths

  • Covers the full lifecycle from data to deployment
  • Three substantial end-to-end projects
  • Practitioner perspective on cost and infrastructure trade-offs

Keep in mind

Fast-moving tooling means some library specifics have already aged, and readers who only build on hosted APIs may find the training and platform chapters more than they need.

Key takeaways

  1. Choosing between API access, fine-tuning an open model and training from scratch is a cost, control and capability trade-off best settled early.
  2. Parameter-efficient methods such as LoRA let teams adapt large models on modest hardware rather than retraining every weight.
  3. Serving LLMs demands dedicated planning for memory, latency and throughput, which differs sharply from deploying conventional machine learning models.
  4. Compact models can run on edge hardware as modest as a Raspberry Pi, widening the places where LLM features can operate.

Notable facts

About the author

Christopher Brousseau and Matthew Sharp

Christopher Brousseau is a staff machine learning engineer at JPMorganChase with a background in linguistics and localization; he has led machine learning and data product initiatives at startups and Fortune 500 companies. Coauthor Matthew Sharp is an engineer, former data scientist and MLOps leader who specializes in deploying, managing and scaling machine learning models in production.

Questions

What does LLMs in Production cover?
LLMs in Production is an engineering guide to building and running applications on large language models. It explains how LLMs work, when to use a hosted model versus training your own, how to build data pipelines and LLMOps platforms, how to fine-tune and serve models, and how to design applications, ending with three hands-on projects including a Raspberry Pi deployment.
Who should read LLMs in Production?
The book targets data scientists and machine learning engineers who know Python and basic cloud deployment and now own an LLM feature or platform. Engineering managers can use its early chapters to frame build-versus-buy decisions. Readers who only want to call a hosted API with prompts will find much of the infrastructure material more detailed than necessary.
Is LLMs in Production current enough for 2026?
Published in late 2024, it covers concerns that still dominate production work, including fine-tuning with LoRA, serving costs, prompt engineering and agents that call external tools. Specific libraries and model versions change quickly, so readers should check current documentation before reusing the code. The architectural reasoning and trade-off discussions remain the book's most durable parts.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.