BrightPool/udemy-prompt-engineering-course

What it is

This is the notebook companion to the Udemy course The Complete Prompt Engineering for AI Bootcamp (2026). It is not a single installable Python package — it's a folder of learning material: Jupyter notebooks, small example apps, prompt files, datasets, screenshots, and diagrams.

At the time of writing, it contains 248 notebooks covering prompt engineering, OpenAI workflows, retrieval, agents, LangChain, LangGraph, evaluation, vision, and image generation.

Who it's for

• Students of the Udemy course who want the hands-on code from the lectures
• Developers looking for practical prompt-engineering and LLM application examples
• Teams that want reference notebooks for OpenAI APIs, RAG, agent patterns, and evaluation workflows

What it covers

The repo is strongest for coding-heavy parts of the curriculum:

• OpenAI API workflows (Responses API, structured outputs, tool calling, code interpreter, computer use)
• Advanced prompting (few-shot, chain-of-thought, self-consistency, prompt optimization)
• Retrieval, embeddings, RAG, hybrid retrieval
• Agents and orchestration (OpenAI Agents SDK, customer support agents, blog generators)
• LangChain and LangGraph deep dives
• Evaluations and optimization (DSPy, SAMMO, eval metrics)
• Vision and image-model projects (multimodal analysis, FLUX/FAL, older Stable Diffusion/DreamBooth)

It also has a section on other providers: OpenRouter, Gemini (video analysis), and Anthropic/Claude.

Note: the repo does not mirror every lecture. Non-coding lessons, platform walkthroughs, and conceptual material live mainly in the videos.

Layout overview

Area	Folders	Contents
Core OpenAI & prompting	openai_features_and_functionality/, advanced_text_model_techniques/	Responses API, tool calling, few-shot, ReAct, self-consistency
Other providers	exploring_other_providers/	OpenRouter gateway, Claude Messages API, Gemini video analysis, provider comparisons
Retrieval & agents	retrieval_embeddings_and_vector_databases/, advanced_retrieval_techniques/, building_ai_agents/, agent_architectures/	Embeddings, hybrid retrieval, support agents, orchestration patterns
Framework dives	deep_dive_on_langchain/, deep_dive_on_langgraph/	LangChain chat models, LCEL, vectorstores; LangGraph state, persistence, human-in-the-loop, RAG
Projects & evals	projects/, prompt_optimization_and_evals/, evaluating_quality/	Blog generation, summarization, transcription, DSPy, SAMMO, eval metrics
Vision & image models	vision/, advanced_image_model_techniques/, ai_image_model_projects/, standard_image_model_practices/	Multimodal analysis, FLUX/FAL workflows, older Stable Diffusion/DreamBooth
Supporting assets	building_ai_agents/resources/, images/, docs/	Sample datasets, prompt files, screenshots, diagrams, internal planning docs

How to run it

Most notebooks are meant for Jupyter or Google Colab. Minimal local setup:

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

Key notes:

• Notebooks install extra dependencies inline with %pip install — there is no single repo-wide requirements.txt or pyproject.toml
• Most notebooks expect an OPENAI_API_KEY
• Many prompt for missing credentials via getpass if not set in the environment

Credentials needed

Mostly just OPENAI_API_KEY. Some sections additionally need:

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

The README explicitly warns: do not commit API keys to git; use environment variables or a secret manager (e.g., 1Password CLI).

Caveats

• deep_dive_on_langgraph/ is a structured sequence, good as a guided track
• deep_dive_on_langchain/ works better as a concept library than a linear path
• building_ai_agents/resources/ holds the datasets, prompts, and knowledge-base files used by agent notebooks
• standard_image_model_practices/ is just a placeholder — not populated
• Several vision/image notebooks depend on external assets, uploads, or cloud credentials, so they aren't fully self-contained after cloning

Suggested learning paths (from README)

1. OpenAI & 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/

For course students: use the videos for the full teaching sequence; treat this repo as the coding companion.

Contributing / maintenance tools

The repo suggests gitleaks (secret scanning), pre-commit, and a gitleaks detect command to keep secrets out of tracked files.