Skip to content

Adnan AlyaariDigital Projects & Transformation

Home Professional Profile Resume Projects Services Contact Learn with Me

Menu

Home Professional Profile Resume Projects Services Contact Learn with Me

Available for work

Language

العربية

Learn with Me/Artificial Intelligence/Prompt Engineering

Current path

Prompt Engineering

Level: Intermediate33 lessons

Progress0/33 · 0%

Lesson index

01Foundations

What Is a Prompt?
What Is Prompt Engineering?
How Models Handle Instructions

02Building Effective Prompts

Define the Task and Goal
Add Context and Inputs
Constraints, Audience, and Tone
Specify the Output Format

03Prompting Techniques

Zero-shot and Few-shot Prompting
Use Examples Effectively
Roles and Personas
Structure Prompts with Delimiters and Sections

04Context and Sources

Working with Texts and Documents
Long-context Prompting
Grounding and Source-based Answers
Missing Information and Hallucinations

05Complex Tasks and Reasoning

Decompose Complex Tasks
Prompt Chaining
Prompting Reasoning Models
When Simpler Instructions Are Better

06Advanced Outputs

Structured Outputs
Information Extraction and Classification
Multimodal Prompting

07Evaluation and Iteration

Why One Run Is Not Enough
Define Success Criteria
Find Failure Cases
Iterative Prompt Improvement

08Practical Applications

Prompts for Research and Learning
Writing, Editing, and Summarization
Information and Data Analysis
Prompts for Work and Productivity
Build a Reusable Prompt

09Capstone Project

From Problem to Final Prompt
Test and Improve the Project

✓Path quiz — 10 questions

Competition is optional

Want to compete for learning titles?

Sign in to save your score and points and optionally join the leaderboard, or continue as a guest.

Final quiz

Test your path knowledge

Answer 10 questions from the path lessons. You need 70% to pass.

1. What is the best step before adding many details to a prompt?

Define the desired outcome and success criteriaMake the prompt as long as possibleAlways add an expert role

2. When is few-shot prompting useful?

When examples clarify the desired pattern or formatOnly when the model cannot readIt must be used in every prompt

3. What is an appropriate use of a role or persona?

Use it when it clarifies perspective, tone, or expertiseAdd it automatically to every promptUse it instead of defining the task

4. Why structure a prompt into clear sections?

Reduce ambiguity between instructions, context, examples, and inputsOnly increase the word countForce the model to reveal internal reasoning

5. When working from supplied documents, what is the better approach?

Require grounding in the supplied material and state when evidence is insufficientFill every gap by guessingIgnore the documents and rely on memory

6. When is prompt chaining useful?

When one stage depends on a previous output and separation improves control or reviewFor every simple questionOnly to change response tone

7. Which principle best fits modern reasoning models?

Define outcomes and constraints clearly and avoid over-prescribing reasoning stepsAlways write the longest possible instruction chainNever specify the output shape

8. Why is one run insufficient to judge prompt quality?

Because outputs can vary and should be tested on representative casesBecause every prompt must produce ten answersBecause testing is unrelated to quality

9. Which is a checkable success criterion?

The output contains all required fields and is grounded in specified sourcesThe answer is amazingThe prompt looks smart

10. What is a good way to iteratively improve a prompt?

Change a specific element, retest, and compare resultsRandomly rewrite everything after each runJudge prompt quality by its length

Adnan AlyaariDigital Projects & Transformation

Ambition begins with an idea... and is completed through achievement

Book a Consultation Contact Share Feedback Policy & Privacy

Home Projects

Learn with Me Services Contact