Prompt Engineering for Professionals | Nesqual TechCourse outline
Prompts at work
- A prompt is a work briefLesson25 min
- What the model does with your textLesson25 min
- The six parts of a work promptLesson30 min
- The course toolkit: Larkfield and the mock modelLesson30 min
- The edit, run, evaluate loopLesson35 min
- What prompting cannot fixLesson30 min
- Chapter quiz: prompts at workQuiz15 min
Writing clear instructions
Specific instructions: the new colleague test
Lesson30 min
Audience, purpose and toneLesson30 minConstraints that work: length, must and must notLesson30 minSeparating instructions from dataLesson30 minOrdering a prompt and asking for no preambleLesson30 minAsk before guessing, and the six-part checkLesson30 minQuiz: writing clear instructionsQuiz15 minLab: four prompts that do exactly what they sayLab40 minContext, grounding and conversations
- Context the model cannot guessLesson30 min
- Grounding answers in supplied materialLesson30 min
- Hallucination in practice: spotting invented factsLesson30 min
- Roles and personas: what "You are..." really doesLesson25 min
- Chat conversations at work: steering, drift and fresh startsLesson30 min
- Working with reference documentsLesson30 min
- Chapter quiz: context, grounding and conversationsQuiz15 min
Examples, style and output formats
- Zero-shot, one-shot and few-shot promptsLesson30 min
- Choosing good examplesLesson30 min
- Style, tone and rewritingLesson30 min
- Output formats for workLesson30 min
- Document skeletons and checking themLesson30 min
- Transformation tasks: summaries, action items and translationLesson30 min
- Quiz: examples, style and output formatsQuiz15 min
Reusable prompt templates with Python
- From a good prompt to a reusable templateLesson25 min
- Python's string.Template: placeholders, substitute and safe_substituteLesson30 min
- f-strings, str.format and why templates live in filesLesson25 min
- Template files with a header blockLesson30 min
- Rendering in batch and running the model from PythonLesson35 min
- Input hygiene for templatesLesson30 min
- Quiz: reusable prompt templatesQuiz15 min
- Lab: a template renderer for the support teamLab40 min
Structured output and validation
- Why structure matters: the nearly right replyLesson25 min
- JSON for non-programmers: reading it and spotting broken JSONLesson30 min
- Asking for JSON that a program can useLesson30 min
- Describing the shape: JSON SchemaLesson30 min
- Writing check_json.py: a validator you can readLesson35 min
- When a reply is invalid: fix the prompt, not the dataLesson25 min
- Chapter quiz: structured output and validationQuiz15 min
- Lab: JSON records the tracker can trustLab40 min
Breaking down complex work
- Why one big prompt fails on multi-step workLesson25 min
- Asking for reasoning: working first, answer lastLesson30 min
- Prompt chains done by handLesson30 min
- Self-review: checklists, critique and reviseLesson30 min
- Options, counter-arguments and stated assumptionsLesson30 min
- When reasoning hurts, and checking the numbersLesson30 min
- Chapter quiz: breaking down complex workQuiz15 min
Tokens, length and context budgets
- Why length matters: tokens at workLesson25 min
- Estimating tokens without a tokeniserLesson30 min
- Prompt budgets and costLesson30 min
- What happens past the windowLesson30 min
- Working with long documentsLesson35 min
- Chapter quiz: tokens, length and context budgetsQuiz15 min
- Lab: count, cut and chunk a long policyLab40 min
Evaluating outputs with rubrics
- Why "it looks good" is not an evaluationLesson25 min
- Defining good: from goals to criteriaLesson25 min
- Analytic rubrics: yes or no criteria versus 1 to 5 scalesLesson30 min
- Automatable and human criteriaLesson30 min
- Scoring in Python: rubric.json and score.pyLesson35 min
- When reviewers disagree: agreement and calibrationLesson30 min
- Quiz: evaluating outputs with rubricsQuiz15 min
- Lab: score four replies with a rubricLab40 min
Iterating, testing and versioning prompts
- One change at a time: the disciplined loopLesson30 min
- A test set and accuracyLesson30 min
- Versions as files and a changelogLesson25 min
- Seeing what changed: diff -u and difflibLesson30 min
- Variation and regressionLesson30 min
- Failure analysis and knowing when to stopLesson30 min
- Quiz: iterating, testing and versioning promptsQuiz15 min
- Lab: test, version and compare a classification promptLab40 min
Safe and responsible use at work
- Confidentiality: what never goes into an assistantLesson30 min
- Redaction before promptingLesson35 min
- Verifying output before you rely on itLesson30 min
- Bias and fairness in workplace promptsLesson30 min
- Untrusted text in your promptsLesson30 min
- Disclosure, ownership and recordsLesson30 min
- Safe and responsible use: chapter quizQuiz15 min
Prompt playbooks and capstone
- Playbooks by role: analysts, writers and editorsLesson30 min
- Playbooks by role: support and engineeringLesson30 min
- A team prompt libraryLesson30 min
- Handing over a promptLesson30 min
- Capstone briefing: the Larkfield triage pipelineLesson30 min
- Capstone lab: the Larkfield triage pipelineLab40 min
- Course review and where to go nextLesson25 min
- Chapter quiz: playbooks, libraries and the capstoneQuiz15 min