AI Infra Handbook - Learning and Practice Notes

In Progress

AI Engineering

AI Infra Handbook is a long-term learning and practice project around AI infrastructure, model serving, vector retrieval, GPU resources, Kubernetes deployment, observability, and cost-aware engineering.

English AI Infra Handbook project cover showing model serving, RAG, Kubernetes, and observability knowledge layers
Created
Updated

Project Profile

Phase
In Progress
Visibility
Learning Handbook
Next Step
Build the map for model serving, RAG, Agent/MCP, GPU, Kubernetes, observability, and cost governance.
Project Relation
The AI Infra knowledge base that complements stage-based blog posts and project reviews.
Boundary
It is a learning and practice handbook, not an authoritative course or unverified production-best-practice claim.
Tech Stack
  • AI Infra
  • LLM
  • GPU
  • Kubernetes
  • Vector Database
  • Observability
AIinfrastructureLLMknowledge-base

One-Line Positioning

AI Infra Handbook is my long-term learning and practice handbook for AI infrastructure. It collects notes on model serving, inference deployment, vector retrieval, GPU resources, Kubernetes, observability, and cost-aware engineering.

It is not an official tutorial or a complete course. It is a living engineering knowledge project that organizes what I study, verify, and practice into reusable notes.

Why It Exists

AI Engineering is more than calling a model API. When systems move toward real delivery, engineers meet problems around model serving, inference performance, context management, vector databases, task orchestration, tracing, cost control, and deployment stability.

This handbook turns those scattered notes into a system that can be revisited and improved over time.

Scope

  • Model serving and OpenAI-compatible API adaptation.
  • RAG, vector databases, embeddings, and retrieval quality.
  • GPU resources, inference services, throughput, latency, and cost.
  • Kubernetes / Docker deployment and runtime observability.
  • Engineering boundaries around Agent, MCP, and Function Calling.
  • Evaluation, regression, logs, traces, and incident reviews.

Current State

The project is in active organization. It first serves my own learning, project reviews, and blog writing. If the structure becomes stable enough, it may be published as a clearer public directory later.

Relationship with the Blog

AI Infra Handbook is the knowledge base for model serving, RAG, agents, deployment, observability, and cost governance. The blog publishes concrete problems, staged practice notes, and project reviews. One is the system directory; the other records the work in progress.

Boundaries

  • No unverified universal architecture claims.
  • Research notes are not presented as production best practices.
  • No internal systems, addresses, secrets, business data, or company-sensitive information are exposed.

Project Asset

Asset Connections

How this project connects to topics, reading paths, series, and future writing.

Related Topics

AI InfraAI Engineering

Reading Paths

Related Series

Related Entries

Project Timeline

  • Map StageBuild the learning map and public boundaries first
  • Writing StageAdd chapters and cases according to the roadmap

Next Content Plan

  • Add LLM architecture, RAG/vector databases, Agent/MCP, model serving and inference deployment.
  • Continue GPU/compute, Kubernetes, evaluation/observability, cost governance, data processing, and engineering cases.

Project FAQ

Is AI Infra Handbook an authoritative course?
No. It is a learning and practice handbook, not a production-best-practice claim.