This repository is the notebook companion for the coding portions of The Complete Prompt Engineering for AI Bootcamp (2026) on Udemy.

It is built around Jupyter notebooks, small example apps, prompt files, datasets, screenshots, and diagrams rather than a single installable Python package. At the time of writing, the repo contains 248 notebooks covering prompt engineering, OpenAI workflows, retrieval, agents, LangChain, LangGraph, evaluation, vision, and image generation.

• Course students who want the code from the hands-on lectures
• Developers who want practical prompt engineering and LLM application examples
• Teams looking for reference notebooks for OpenAI APIs, RAG, agent patterns, and eval workflows

This repository does not mirror every lecture one-to-one.

The full course includes non-coding lessons, platform walkthroughs, and conceptual material that live mainly in video. The repo is strongest for the coding-heavy sections of the curriculum, especially:

• OpenAI API workflows
• Advanced prompting techniques
• Retrieval, embeddings, and RAG
• Agent design and orchestration
• LangChain and LangGraph
• Prompt optimization and evaluations
• Vision and image-model projects

If you are new to the repo, pick a track below:

• OpenAI Playground: browser introduction and links
• Introduction to the Responses API
• Chat Completions vs Responses API
• Responses API & Messages
• Structured Outputs
• Decisions API: typed answers, speed and cost
• Tool Calling
• Code Interpreter: Python in a Sandbox
• Computer Use Through the API
• Managed Agents: The Agents API

• Section guide: lesson order and comparison notes
• OpenRouter Features and Functionality
• A/B Test the Model, Keep the Prompt
• Using Gemini for Video Analysis, native SDK, timestamped Q&A and structured outputs
• Building a Small Agent with OpenRouter
• Anthropic Features and Functionality
• Claude Prompting and Python Tools
• Claude Computer Use, optional live demo disabled by default

• Few-Shot Learning
• Chain of Thought
• Self-Consistency Sampling
• Prompt Optimization

• Introduction to Retrieval
• Understanding Vector Embeddings
• RAG with Vector Databases
• Hybrid Retrieval with SQL Filtering & Embeddings

• Multi-Source Customer Support Agent
• Introduction to OpenAI Agents SDK
• Skills Based Agents
• Blog Post Generator Agent

• Simple LangGraph through Time Travel

• Social Media Posts
• Evaluation Metrics
• DSPy Primer

Most notebooks are designed to be opened in Jupyter or Google Colab.

For local use, a minimal setup looks like this:

python -m venv .venv
source .venv/bin/activate
pip install jupyterlab
jupyter lab

Notes:

• Many notebooks install additional dependencies inline with %pip install
• There is no single repo-wide requirements.txt or pyproject.toml for every notebook
• Most notebooks expect OPENAI_API_KEY
• Many notebooks will prompt for missing credentials with getpass if they are not already set in your environment

Do not commit API keys or secrets to git.

Most notebooks can be run with just OPENAI_API_KEY, but some sections need extra services:

• ANTHROPIC_API_KEY for the Claude API section and prompt-caching comparisons
• SUPABASE_URL and SUPABASE_KEY for pgvector examples
• LANGCHAIN_API_KEY and TAVILY_API_KEY for some agent and tracing workflows
• FAL_KEY for several FLUX image-generation notebooks
• Hugging Face or Google Cloud credentials for some older image-model project notebooks

Use environment variables or your preferred local secret manager. If you use 1Password CLI, inject secrets locally rather than storing them in tracked files.

Some folders are more complete than others:

• deep_dive_on_langgraph/ is a structured sequence and works well as a guided learning track
• deep_dive_on_langchain/ is richer as a library of concepts than as a strict linear path
• building_ai_agents/resources/ contains the datasets, prompts, and knowledge-base files used by the agent notebooks
• standard_image_model_practices/ is currently just a placeholder and should not be treated as a populated section
• Several vision and image notebooks depend on external assets, uploads, or cloud credentials and are not fully self-contained after clone

If you are maintaining or updating the repo, the existing hygiene tools are still useful:

brew install gitleaks
pip install pre-commit
pre-commit install
gitleaks detect --source="." --report-path="gitleaks-report.json"

If you want a more deliberate progression:

1. OpenAI and prompting: openai_features_and_functionality/ -> advanced_text_model_techniques/
2. Retrieval: retrieval_embeddings_and_vector_databases/ -> advanced_retrieval_techniques/
3. Agents: building_ai_agents/ -> agent_architectures/
4. Frameworks: deep_dive_on_langchain/ and deep_dive_on_langgraph/
5. Applied projects: projects/, prompt_optimization_and_evals/, vision/, advanced_image_model_techniques/

If you are following the Udemy course, use the videos for the full teaching sequence and use this repo as the coding companion.