References

A curated collection of resources for deepening your understanding of agentic AI patterns.

OpenAI

• Chat Completions API Guide — Primary API used in this workshop
• Function Calling Guide — Tool definitions and usage
• Structured Outputs Guide — response_format with Pydantic
• Responses API Guide — OpenAI's new primary API
• API Reference — Complete API documentation
• OpenAI Cookbook — Practical examples and recipes
• Tokenizer Tool — Visualize how text is tokenized
• Model Overview — Available models and capabilities

Microsoft / Azure

• AI Agent Design Patterns (MS Learn) — The primary reference for orchestration patterns in this workshop
• Azure OpenAI Service Documentation — Azure-hosted OpenAI models
• Azure AI Foundry — Model deployment and management
• Microsoft Agent Framework — Microsoft's multi-language SDK (Python + .NET) for building AI agents and multi-agent workflows; supports sequential, concurrent, group chat, and handoff patterns with graph-based orchestration and MCP tool integration
• Azure AI Agent Service — Managed service for building and deploying AI agents at scale in Azure AI Foundry

Anthropic / Claude

• "Building Effective Agents" — Excellent overview of agent patterns including evaluator-optimizer and tool use
• Prompt Engineering Guide — Best practices for writing effective prompts

Blog Posts & Articles

• Lilian Weng — "LLM Powered Autonomous Agents" — Comprehensive survey of agent architectures, planning, memory, and tool use
• Chip Huyen — "Building a Generative AI Platform" — Production considerations for AI systems

YouTube / Video

• Andrej Karpathy — "Intro to Large Language Models" — Foundational understanding of how LLMs work
• Andrew Ng — "What's next for AI agentic workflows" — Overview of agentic design patterns (reflection, tool use, planning, multi-agent)
• Sam Witteveen — Multi-Agent Systems — Practical tutorials on building multi-agent systems

Research Papers

• ReAct: Synergizing Reasoning and Acting in Language Models (Yao et al., 2023) — The foundational Reason-Act-Observe loop used in our agent abstraction
• Reflexion: Language Agents with Verbal Reinforcement Learning (Shinn et al., 2023) — Self-reflection and self-evaluation patterns; basis for the maker-checker exercise
• CAMEL: Communicative Agents for "Mind" Exploration (Li et al., 2023) — Multi-agent collaboration and communication protocols
• Toolformer: Language Models Can Teach Themselves to Use Tools (Schick et al., 2023) — How models learn to use tools effectively
• Chain-of-Thought Prompting (Wei et al., 2022) — Step-by-step reasoning that improves complex task performance

Python Libraries Used

• OpenAI Python SDK — pip install openai
• Pydantic — Data validation and settings management
• python-dotenv — .env file loading
• MKDocs Material — Documentation theme

Frameworks (for Reference)

These frameworks implement the patterns taught in this workshop. Understanding the patterns from scratch makes you a more effective user of any of these:

• Microsoft Agent Framework — Microsoft's multi-language SDK (Python + .NET) for building AI agents and multi-agent workflows. Supports sequential, concurrent, group chat, and handoff patterns with graph-based orchestration, session state management, and MCP tool integration
• LangGraph — Graph-based agent orchestration (state machines, conditional routing)
• Pydantic AI — Type-safe agent framework built on Pydantic
• CrewAI — Role-based multi-agent orchestration