Prompt Engineering

Prompt Engineering on Breni picks up where general AI knowledge stops: it teaches how to write instructions that get accurate, usable work from models such as ChatGPT, Claude and Gemini. Across 7 courses it moves from how models read prompts to clear instructions, examples and context, reasoning prompts, structured output, testing and real work tasks. Learners finish able to build prompt chains, get dependable JSON back and compare prompt versions on a test set. Testing opens only after structured output, as courses unlock one at a time.

• 7 courses
• 10 learning objectives
• Level: beginner
• Language: English
• Price: Subscription, free trial

What you will be able to do

• Write prompts that set a clear task, audience, format and length
• Use few-shot examples and delimiters to control how a model responds
• Break a complex task into a chain of smaller prompts
• Get reliable JSON or table output from a chatbot
• Test two prompt versions on the same inputs and keep the better one
• Recognise prompt injection and made-up facts before relying on an answer

Full curriculum

Part 1 · Prompt basics

1. How Language Models Read Prompts: A Beginner Guide

Explain how a model turns a prompt into a reply and why wording matters. Units, in order: Tokens and the context window; Why models predict rather than look things up; System, user and assistant messages; Temperature and randomness; Common ways prompts fail.

• — Tokens and Context Window
  Learn how models break text into tokens and manage working memory.
  • What is a token?
  • The context window limit
• — Prediction Over Lookup
  Understand why models predict next words instead of searching files.
  • Models predict next words
  • Memory without lookup
• — System and User Messages
  Guide AI behavior using system, user, and assistant message types.
  • The system prompt
  • User and assistant roles
• — Temperature and Randomness
  Control model creativity and consistency using the temperature setting.
  • What temperature does
  • Low versus high settings
• — Common Prompt Failures
  Spot why prompts fail and learn how to fix them.
  • Vague instructions
  • Overloading the model

2. Writing Clear Prompts

Write prompts that state the task, audience and format without ambiguity. Units, in order: Stating the task and the goal; Giving the model a role and an audience; Specifying length and format; Positive instructions over prohibitions; Iterating on a weak prompt.

3. Examples and Context

Use examples and background material to steer a model's answers. Units, in order: Zero-shot and few-shot prompting; Choosing good examples; Pasting documents as context; Delimiters to separate instructions from data; Long context and what gets missed.

Part 2 · Advanced techniques

4. Reasoning Prompts

Get models to work through multi-step problems more reliably. Units, in order: Asking for step-by-step reasoning; Breaking a task into smaller prompts; Prompt chaining; Self-checking and critique prompts; When reasoning prompts do not help.

5. Structured Output

Get answers back in a fixed format that other tools can use. Units, in order: Tables and bullet formats; JSON output and schemas; Extracting fields from messy text; Classifying text into categories; Validating what comes back.

Part 3 · Prompts in practice

6. Testing and Improving Prompts

Measure whether a prompt works across many inputs, not just one. Units, in order: Building a small test set; Comparing two prompt versions; Spotting made-up facts; Grounding answers in sources; Keeping a prompt library.

7. Prompting for Real Work

Design prompt workflows for writing, research, coding and data tasks. Units, in order: Drafting and editing writing; Research and summarising; Writing and explaining code; Image generation prompts; Prompt injection and safe use.

Sample: How Language Models Read Prompts: A Beginner Guide

The opening tier, (Tokens and Context Window). Learn how models break text into tokens and manage working memory.

What is a token?

Models read text as small chunks called tokens, not whole words.

The context window limit

The context window is text the model remembers at once.

Questions about Prompt Engineering

What is prompt engineering?

Prompt engineering is the skill of writing instructions that get accurate, useful output from AI models. It covers wording, examples, context, output format and testing.

Do I need coding skills for prompt engineering?

No. Everything in the path works in a chat window, though the structured output course shows how prompts can feed other tools.

Is prompt engineering still worth learning?

Yes, for anyone who uses AI tools often, because clear instructions and careful testing still change the quality of answers a lot. Breni teaches habits that carry across models rather than tricks for one chatbot.

How is this different from AI for Beginners?

AI for Beginners explains what AI is and how it works. Prompt Engineering assumes you already use AI tools and focuses on getting better results from them.

Do I get a certificate in prompt engineering?

Finishing all 7 courses earns a Breni certificate of completion. It records that you completed the path and is not a professional qualification.

Related learning paths

• C++ for Beginners
• Java for Beginners
• Cyber Security for Beginners
• Cloud Computing and AWS
• UI/UX Design for Beginners
• Data Analysis for Beginners
• React for Beginners
• AI for Beginners
• Learn to Code

Start Prompt Engineering

Get Breni on Google Play and start Prompt Engineering Download on the App Store