Azure OpenAI + LLM Wiki

A comprehensive, curated collection of resources for Azure OpenAI, Large Language Models (LLMs), and their applications.

🔹Concise Summaries: Each resource is briefly described for quick understanding
🔹Chronological Organization: Resources appended with date (first commit, publication, or paper release)
🔹Monthly Updates: The list is updated monthly; candidate entries before the update are tracked in the issue.

🧭 Quick Navigation (Propedia-style)

Layer / Era	What it controls	Jump to sections
Weights 2022-2023	Parametric knowledge baked into the model. Themes: Pretraining, Scaling Laws, Fine-tuning, RLHF, Alignment, Instruction-following, Few-shot	Foundations: Large Language Model Landscape, Large Language Model Collection, Foundation Model Providers Training: Large Language Model Training and Optimization, Model Training & Inference, Training & Fine-tuning Behavior and safety: Trust, Safety, and Security, Safety, Security & LLMOps
Context 2023-2024	What the model sees at inference time. Themes: Prompting, Chain-of-Thought, RAG, Memory, Long Context, Knowledge Injection, Context Engineering	Prompting: Prompt Engineering and Visual Prompts, Prompt Engineering & Tooling Retrieval: RAG, Azure AI Search, RAG Best Practices Memory and context windows: Context and Long-Context Limits, Memory, Data Processing & Memory
Harness 2025-2026	How the agent acts in the real world. Themes: Function Calling, Tool Ecosystems, MCP, Skills, Workflow Graphs, Multi-agent, A2A protocols, Orchestration, Agent Infrastructure, Security	Agent runtime: AI Application, Agent Frameworks, Agent Development, Agent Best Practices Protocols and tools: Agent Protocol, Coding & Research, Skills, Harness, Loop Engineering, Dev Tools, MCP & Extensions Apps and operations: Evaluating Large Language Models, LLMOps, Learning Resources & Workshops, Code Samples & Workshops

Refereces: DailyDoseOfDS - Evolution of the Agent Landscape

1. App & Agent

🚀 RAG Systems, LLM Applications, Agents, Frameworks & Orchestration

• RAG
  • RAG
  • GraphRAG
  • RAG Application
  • Vector Database & Embedding
• Application
  • Top Agent Frameworks
  • Additional Agent Framework
  • Cache
  • Data & Analytics Agents
  • Data Processing & OCR
  • Desktop AI Assistant
  • Memory
  • Model Gateway
  • Model Serving & Local Runtimes
  • Observability & LLMOps
  • Popular LLM Applications (GitHub Stars >= 1000)
  • SDKs, Integration & ML Libraries
  • Training & Fine-tuning
  • UI & No-Code Tool
• Agent Protocols
  • A2A
  • Computer use
  • Model Context Protocol (MCP)
• Coding & Research
  • Coding
  • Deep Research
  • Domain-Specific Agents
  • Skills
  • Harness
    • Loop Engineering

⬆ back to top

2. Azure OpenAI & Copilot

🌌 Microsoft's Cloud-Based AI Platform and Services

• Overview
  • Azure OpenAI & Foundry Overview
• Frameworks
  • Orchestration Frameworks
  • Agent Frameworks
• Tooling
  • Prompt Engineering & Tooling
  • Dev Tools, MCP & Extensions
• Products
  • Copilot Product Catalog
  • Agent Development
  • Microsoft 365 Agent Development
• Services
  • Azure AI Search
  • Microsoft Foundry & AI Services
• Research
  • Microsoft Research
• Applications
  • Sample Applications
  • Solution Accelerators
  • Architecture Patterns & Use Cases

⬆ back to top

3. Research & Survey

🧠 LLM Landscape, Prompt Engineering, Finetuning, Challenges & Surveys

• Landscape
  • Large Language Model Landscape
  • Large Language Model Collection
  • Foundation Model Providers
  • Domain-Specific and Specialized LLMs
• Prompting
  • Prompt Engineering and Visual Prompts
• Training & Optimization
  • Large Language Model Training and Optimization
  • Pre-training and Data Preparation
  • Architecture and Inference Patterns
  • Post-training and Fine-Tuning
• Impact & Products
  • AI Adoption, Impact, and Society
  • OpenAI Products
  • Anthropic AI Products
  • Google AI Products
• Survey & Reference
  • Survey and Reference
  • Additional Topics: A Survey of LLMs
  • LLM Research (Ranked by cite count ≥150)
  • Learning Resources, Implementations, and Regional Materials

⬆ back to top

4. Datasets, Evaluation, and Extras

🛠️ Training Data, Datasets & Evaluation Methods

• Data
  • Datasets for LLM Training
• Evaluation
  • Evaluating Large Language Models
  • LLM Evaluation Benchmarks
  • LLMOps: Large Language Model Operations
• Extras
  • LLM for Robotics
  • Awesome Demo

⬆ back to top

5. Best Practices

📋 Curated Blogs, Patterns, and Implementation Guidelines

• RAG
  • The Problem with RAG
  • RAG Solution Design
  • RAG Research
  • RAG Research (Ranked by cite count >=100)
• Agent
  • Agent Design Patterns
  • Agent Research
  • Agent Research (Ranked by cite count >=100)
  • Tool Use: LLM to Master APIs
• Security
  • Security and Governance
• Reference
  • Proposals & Glossary

⬆ back to top

🧭 Start Here

Category	Goal	Suggested path
RAG	Explore RAG patterns	RAG → GraphRAG → RAG Application → RAG Best Practices → RAG Research
AI Engineering	Build an AI Engineering Workflow	RAG → AI Application → Agent Protocol → Coding → Deep Research → Domain-Specific Agents → Skills → Harness → Loop Engineering
Skills & Harnesses	Extend a Coding Agent	Skills → Harness → Loop Engineering → Coding → Tool Use → Evaluation Metrics
Agents	Design an agent workflow	Top Agent Frameworks → Agent Design Patterns → Tool Use → Memory → Agent Research
Data & Analytics	Build a data or analytics agent	Data & Analytics Agents → Data Processing & OCR → Memory → Tool Use → Evaluating Large Language Models
Local LLMs	Build a local or self-hosted LLM application	Large Language Model Collection → Model Serving & Local Runtimes → Model Gateway → UI & No-Code Tool → Observability & LLMOps
MCP & Integration	Build MCP-enabled tools	Model Context Protocol → Dev Tools, MCP & Extensions → Safety, Security & LLMOps → Agent Best Practices
Developer Agents	Build coding or research agents	Coding → Deep Research → Skills → Harness → Loop Engineering → Tool Calling & Agentic
Azure / RAG	Build an Azure RAG application	Azure OpenAI & Foundry Overview → Azure AI Search → RAG Solution Design → Sample Applications → Evaluating Large Language Models
Azure / Agents	Build an Azure agent	Agent Frameworks → Agent Design Patterns → Model Context Protocol → Agent Development → Evaluating Large Language Models
Microsoft 365	Build a Microsoft 365 agent	Microsoft 365 Agent Development → Copilot Product Catalog → Dev Tools, MCP & Extensions → Agent Development
Production	Operate an AI application in production	Architecture Patterns & Use Cases → Safety, Security & LLMOps → LLMOps → Evaluating Large Language Models
Research	Learn the LLM landscape	Large Language Model Landscape → Survey and Reference → LLM Research
Model Development	Train or fine-tune a model	Large Language Model Collection → Model Training & Inference → Training & Fine-tuning → Datasets for LLM Training → Evaluating Large Language Models
Multimodal	Build a multimodal application	Multimodal Models → Data Processing & OCR → RAG Application → Vision & Multimodal
Evaluation	Choose and benchmark a model	Large Language Model Collection → Architecture Comparisons → Evaluating Large Language Models → LLM Evaluation Benchmarks → Evaluation Metrics

📖 Legend & Notation

Symbol	Meaning	Symbol	Meaning
	GitHub repository	🗄️	Archived files
💡🏆	Recommend	📺	Video content
📑	Academic paper	🤗	Huggingface

Info: Applications that have been archived or have had no commits for more than 12 months are listed in applications.old.md.

^ back to top ^