Platform Engineering for Artificial Intelligence: Mastering Kubernetes Orchestration for Natural Language Processing with LLM - Couverture souple

Stormveld, Erik

Couverture souple

ISBN 13 : 9798171899660

Editeur : Independently published, 2026

Synopsis

What happens when AI experimentation needs to become reliable enterprise infrastructure?

Building an LLM application is one challenge. Building the platform that allows AI teams to deploy, scale, secure, observe, and continuously improve those applications is another.
Platform Engineering for Artificial Intelligence: Mastering Kubernetes Orchestration for Natural Language Processing with LLM explores the architecture behind modern enterprise AI platforms, bringing together Kubernetes, large language models, RAG, vector databases, GPUs, agentic workflows, and LLMOps.
Designed for engineers and architects working at the intersection of cloud infrastructure and AI, this book explores how to create standardized, scalable foundations for production LLM workloads.
Inside, you'll explore:

• Designing internal AI platforms and developer-friendly “golden paths”
• Deploying and serving LLMs with vLLM and TGI on Kubernetes
• Scaling vector databases and persistent embedding infrastructure
• Building and operating RAG pipelines with event-driven workloads
• Orchestrating MCP gateways, tools, and multi-agent systems
• Running LoRA, QLoRA, and distributed fine-tuning workflows
• Managing LLM gateways, token budgets, caching, and model routing
• Implementing security, guardrails, RBAC, PII protection, and compliance
• Establishing observability and LLMOps across the model lifecycle

If you're ready to move beyond isolated AI prototypes and understand the infrastructure required to support scalable, governed, production-ready enterprise AI, this book provides a practical architectural foundation for the journey.
Build the platform your AI teams can depend on.

Détails de l'édition

• Éditeur: Independently published
• Date d'édition: 2026
• Langue: anglais
• ISBN 13: 9798171899660
• Reliure: Paperback
• Nombre de pages: 188
• Coordonnées du fabricant: non disponible
• Personne responsable: non disponible