Prompt Engineering for Everyone Bootcamp

Discover prompt engineering, the art and science of crafting inputs for large language models to define context, role, and format for useful, accurate, and safe outputs.

Advance organizational impact by learning prompt engineering across industries; it bridges raw AI to real-world utility, democratizing AI for healthcare, education, software, marketing, and data science.

Explore the ethical use of AI and how prompt engineers shape its real-world impact. Learn to reduce hallucinations, verify outputs, prevent bias, and protect privacy in health, law, and education.

Explore zero-shot prompting, a simple method to interact with large language models by giving instructions without examples, and learn key prompt engineering terms used for tasks like summarization and translation.

One-shot prompting uses a single example to convey the desired pattern, improving reliability for subjective tasks like translation and sentiment analysis by defining the task and showing one input-output.

Few-shot prompting uses two or more examples to prime the model, improving accuracy for classification, extraction, and formatting. It shows patterns that guide tone and logic for consistent results.

Explore chain-of-thought prompting by asking models to show their work step by step, using medical and math examples to reveal reasoning and improve problem solving.

Master iterative prompting by starting with initial prompts, evaluating outputs, refining prompts through feedback and rephrasing, and repeating until satisfied, with examples from climate change and code debugging.

Explore negative prompting as a core prompt engineering technique to prevent hallucinations, control tone, and restrict style by specifying what not to do.

Discover how role-based prompting guides large language models to adopt specific personas and tones for audience-aware responses. Follow a simple workflow: add a role, include context, and refine the prompt.

Hybrid prompting blends multiple prompting techniques—examples, reasoning steps, constraints, and iterative and negative prompts—to improve accuracy, control, and depth for complex tasks and ambiguity.

Develop a professional, empathetic response to a delayed delivery without blaming the carrier. Learn to hybridize role prompting, negative prompting, and one-shot prompting to craft effective prompts.

Learn how language models generate text token by token, predicting the next token by probabilities, with tokens defined as words or parts of words, and how prompt accuracy shapes responses.

Discover what tokens are in large language models using the OpenAPI tokenizer, see how text is split into tokens rather than words, and learn how tokens impact cost and performance.

Explore popular large language models, their organizations and access methods, from GPT and Claude to llama, mistral, and Gemini, plus Amazon Bedrock and Titan embedding.

Explore how transformers power large language models by detailing embedding, positional encoding, encoder-decoder layers, and attention mechanisms. Learn how self-attention, multi-head attention, feedforward networks, and softmax produce translations and outputs.

Explore how transformers power models like GPT-4, Claude, and Gemini, and learn how attention, tokens, and semantic similarity shape prompts to improve few-shot and chain-of-thought prompting.

Craft clear prompts by specifying constraints and output formats to avoid vague, off-target responses. Set audience, length, and format, such as three historical facts for kids in 100 words.

Learn how to craft effective prompts by adding context and background—defining roles, audience, data, and task—to guide responses and achieve precise, audience-appropriate summaries.

Practice crafting clear prompts for large language models to explain photosynthesis to teens aged 13 to 17, with audience, tone, and format in mind.

Practice iterative prompt testing: start simple, test, and refine with constraints until output matches expectations, especially for legal, medical, or product descriptions.

Learn to tune prompts and parameters for image generation with Python and the OpenAI image model, including handling API keys, image size, quality, and style settings.

Learn to set up the llama 2 model, install and configure it, and send API requests with a structured payload (model, prompt, temperature, top_p, top_k, max tokens) to generate text.

Meta prompting asks the AI to create, refine, or evaluate prompts, helping you structure complex tasks, improve prompt quality and tone, and build reusable templates.

Explore meta prompting for text generation with large language models, learning to design role based prompts that improve tone, detail, and outcomes across tasks like emails and summaries.

Learn meta prompting for code generation, enabling non-programmers to create a beginner-friendly command-line to-do list app in Python via role-based prompts and a detailed meta prompt.

Explore meta prompting to boost image generation by crafting role-based prompts and applying them in the same or different models to create cinematic futuristic city at night visuals.

Learn what AI hallucination is and how prompting reduces its impact in large language models, with care in high-stakes settings and the need for grounding via retrieval tools.

Learn prompting techniques to reduce hallucination by guiding models with role prompts, constraints, and citations, and by using guardrails, verification, and negative prompting.

Learn to craft prompts that minimize AI hallucination by requiring citations, requesting known works, and limiting speculation to current knowledge (as of 2025) across models.