AI Production Engineering Handbook
How AI systems are built, run and governed in production. 18 books · 1,005 pages
i
About this collection
A capability inventory for production AI applications, agent systems, and enterprise platforms. Each capability's README.md states neutral, durable guidance — concept, practices,…Open the book
About this collectionA capability inventory for production AI applications, agent systems, and enterprise platforms. Each capability's README.md states neutral, durable guidance — concept, practices,…
Shared
Shared
01
Model Management & Inference
Scope: model selection, serving, and inference.Open the book
Model Management & InferenceScope: model selection, serving, and inference.
02
Agent Architecture & Orchestration
Scope: agent planning, coordination, execution, and the systems that make agent work reliable.Open the book
Agent Architecture & OrchestrationScope: agent planning, coordination, execution, and the systems that make agent work reliable.
03
Tools, MCP & External Integration
Scope: external system access and side-effectful actions.Open the book
Tools, MCP & External IntegrationScope: external system access and side-effectful actions.
04
Memory & State Management
Scope: durable and scoped recollection of prior context.Open the book
Memory & State ManagementScope: durable and scoped recollection of prior context.
05
Context Engineering & Prompt Management
Scope: constructing effective, secure model context.Open the book
Context Engineering & Prompt ManagementScope: constructing effective, secure model context.
06
RAG & Knowledge Engineering
Scope: retrieval and enterprise knowledge grounding.Open the book
RAG & Knowledge EngineeringScope: retrieval and enterprise knowledge grounding.
07
Guardrails & AI Safety
Scope: runtime safeguards against unsafe input, output, and actions.Open the book
Guardrails & AI SafetyScope: runtime safeguards against unsafe input, output, and actions.
08
Evaluation & Quality Engineering
Scope: measurement of application quality and failure modes.Open the book
Evaluation & Quality EngineeringScope: measurement of application quality and failure modes.
09
Observability, Tracing & Monitoring
Scope: inspectable operations and diagnostic evidence.Open the book
Observability, Tracing & MonitoringScope: inspectable operations and diagnostic evidence.
10
Security, Identity & Access Control
Scope: identity boundaries, confidentiality, and authorized actions.Open the book
Security, Identity & Access ControlScope: identity boundaries, confidentiality, and authorized actions.
11
Workflow, Execution & Reliability Engineering
Scope: durable automation with recoverable side effects.Open the book
Workflow, Execution & Reliability EngineeringScope: durable automation with recoverable side effects.
12
Human–AI Interaction & Application UX
Scope: safe, clear human interaction with AI workflows.Open the book
Human–AI Interaction & Application UXScope: safe, clear human interaction with AI workflows.
13
AI Governance & Compliance
Scope: organizational accountability, policy, and risk.Open the book
AI Governance & ComplianceScope: organizational accountability, policy, and risk.
14
Cost, Performance & FinOps
Scope: measuring and optimizing unit economics.Open the book
Cost, Performance & FinOpsScope: measuring and optimizing unit economics.
15
Infrastructure, Deployment & MLOps/LLMOps
Scope: release engineering, infrastructure, and service resilience.Open the book
Infrastructure, Deployment & MLOps/LLMOpsScope: release engineering, infrastructure, and service resilience.
16
Platform Engineering & Developer Experience
Scope: reusable enterprise developer tooling and self-service.Open the book
Platform Engineering & Developer ExperienceScope: reusable enterprise developer tooling and self-service.