01 / Course at a closer look
Course overview
Learn AI fundamentals, write effective prompts and build practical workflows for communication, content and data tasks in a beginner-friendly 30-hour course.
Build a practical understanding of artificial intelligence and learn how to use it thoughtfully in work and study. This beginner-friendly course explains AI, machine learning and generative AI in plain language before introducing hands-on techniques for prompting, communication, creative planning, structured data extraction and workflow design.
Across ten modules, you will practise using public or fictional information, compare outputs with their source material and improve prompts using clear quality criteria. You will also learn the basics of model evaluation, recognize misleading results and plan safeguards for privacy, bias and unreliable outputs.
The course concludes with a small capstone project: design an AI-assisted solution, test it against realistic and difficult inputs, document its limitations and present the result. Coding is optional; the focus is on understanding, practical application and human review.
Proposed learning format: 30 hours of guided lessons, demonstrations, exercises and project work. Assessment uses module knowledge checks, practical deliverables and the final project. Learners should have access to a computer, a spreadsheet application and an institution-approved AI assistant. Tool availability and any paid account requirements must be confirmed by the training provider.
What makes up the program?
- Understanding Artificial Intelligence (2 hours)
- How Generative AI and Language Models Work (3 hours)
- Prompt Design and Iterative Improvement (4 hours)
- AI for Communication, Research and Learning (3 hours)
- AI-Assisted Content and Creative Planning (3 hours)
- Working with Data and Structured Outputs (3 hours)
- Machine Learning and Model Evaluation (3 hours)
- Designing Reliable AI Workflows (3 hours)
Who is this course for?
Students, graduates, working professionals, educators, administrative teams, content creators and small-business teams seeking a practical introduction to AI. Suitable for learners without a programming background.
02 / Skills you'll build
What you'll learn
01
Distinguish AI, machine learning, deep learning and generative AI.
02
Explain basic language-model behaviour and recognize unsupported outputs.
03
Write and improve prompts using task, context, constraints and output format.
04
Draft and review emails, summaries, study plans and creative briefs.
05
Extract structured records and verify completeness, data types and totals.
06
Distinguish classification, regression and clustering and interpret basic evaluation measures.
07
Design AI-assisted workflows with validation, exceptions and human approval.
08
Identify privacy, bias, security and reliability risks and propose controls.
09
Build, test and present a small AI-assisted capstone solution.
03 / Module breakdown
Detailed course curriculum
01Understanding Artificial Intelligence
- Learning objective: Explain what AI does and identify suitable everyday applications.
- Lesson content: Artificial intelligence describes systems designed to perform tasks such as recognizing patterns, interpreting language, making predictions and generating content. Machine learning learns patterns from examples; deep learning uses multi-layer neural networks. Generative AI creates new outputs such as text or images. A rule-based system follows explicitly written instructions. These approaches overlap, but they are not interchangeable.
- Examples: Email spam detection is classification; estimating delivery time is prediction; drafting an email is generation. An AI system can be useful without understanding a task as a person does. Its output depends on its data, design and input.
- Practice: List five tasks from work or study. Classify each as rule-based automation, prediction, generation or a task needing human judgement. Select one task with a clear, measurable outcome.
- Deliverable: A one-page AI opportunity map containing task, input, expected output and human reviewer.
- Knowledge check: Distinguish AI, machine learning and generative AI using your own examples.
02How Generative AI and Language Models Work
- Learning objective: Describe language-model behaviour and recognize its practical limits.
- Lesson content: A language model generates text by predicting likely next tokens from the context available to it. Tokens are pieces of text; they are not always whole words. Training shapes model behaviour, while inference produces a response to a new input. A context window limits the information a model can consider at once. A fluent answer can still contain invented facts, incorrect calculations or unsupported sources.
- Explore the difference between asking a model from memory, providing a document as context and retrieving information from an external source. Tool access, file handling and image understanding depend on the system being used. Do not assume that an assistant has searched the web or read a file unless that action is supported and confirmed.
- Practice: Ask a generic factual question, then repeat it with a short supplied source. Compare specificity, evidence and unsupported claims. Use public or fictional information only.
- Deliverable: A comparison table with response, supporting evidence, uncertainty and correction.
- Knowledge check: Explain why confidence and fluent writing do not prove accuracy.
03Prompt Design and Iterative Improvement
- Learning objective: Write clear prompts and improve them using an observable quality test.
- Lesson content: A useful prompt states the task, audience, context, constraints and output format. Give examples when the desired structure is hard to describe. Separate source material from instructions. Specify how missing information should be handled. Request a concise explanation and supporting evidence when needed; detailed reasoning text is not a substitute for verification.
- Prompt example: "Using only the notes below, write a 120-word course announcement for beginners. Include the learning goal, three topics and a next step. Do not invent a price, date or trainer. If a required detail is absent, mark it as missing. Notes: [insert notes]."
- Practice: Start with a vague prompt for an email or summary. Improve it in three iterations by adding audience, constraints and an example output. Score each result for relevance, completeness, factual support and format compliance.
- Deliverable: A reusable prompt library with five templates: drafting, summarizing, extracting, comparing and planning. Include one test case for each template.
- Knowledge check: Rewrite a prompt that contains ambiguous instructions and conflicting requirements.
04AI for Communication, Research and Learning
- Learning objective: Use AI to draft and organize information while preserving evidence and meaning.
- Lesson content: Apply AI to emails, meeting notes, study plans, plain-language explanations and document summaries. Define the audience and tone before drafting. A summary should preserve important qualifications, decisions and unanswered questions. For research, separate source discovery from evidence review. Check that a cited source exists and actually supports the claim; a generated citation can be fabricated. Keep an evidence log for claims that matter.
- Prompt example: "Summarize these meeting notes into decisions, action items and open questions. Include an owner or due date only when the notes state one. Flag missing details."
- Practice: Use a fictional meeting transcript to produce a summary and follow-up email. Trace every action item to the original notes. Create a study plan for a topic using a provided syllabus, then verify its coverage.
- Deliverable: A reviewed meeting summary, email and source-checking checklist.
- Knowledge check: Identify where a summary changes the meaning of its source and correct it.
05AI-Assisted Content and Creative Planning
- Learning objective: Produce a consistent content brief and evaluate creative outputs against it.
- Lesson content: Build a creative brief containing audience, purpose, message, tone, channel and constraints. Use AI to explore headlines, course descriptions, social posts, image concepts and storyboard ideas. Compare alternatives against the brief rather than choosing the most polished wording. Verify factual claims, remove unsupported promises and review accessibility. Keep brand names, contact information and numerical claims consistent.
- Image-prompt planning can specify subject, composition, style, lighting and intended use. Generated images may contain incorrect text or distorted details, so inspect outputs before reuse. Rights and permitted use depend on source material and tool terms; check these before publishing.
- Practice: Create a fictional training campaign with a course introduction, three short posts and a six-frame storyboard. Revise exaggerated claims and write descriptive alternative text for the proposed images.
- Deliverable: A small content pack and a checklist covering consistency, evidence, accessibility and approval.
- Knowledge check: Explain why a visually convincing output still requires content review.
06Working with Data and Structured Outputs
- Learning objective: Extract structured information and verify it against source records.
- Lesson content: Distinguish text, numbers, dates, categories and identifiers. Missing values are different from zero. Define an output schema before asking AI to extract records: field names, data types, required fields and rules for unavailable information. AI can help explain spreadsheet formulas, categorize text and draft tables, but its arithmetic and row extraction must be checked. A generated table is not evidence that every source row was captured.
- Prompt example: "Extract item_name, quantity and unit_price from the supplied fictional invoice. Return a table with one row per item. Use null for unavailable values. Preserve item names and identifiers exactly. Do not guess unreadable text."
- Practice: Extract ten fictional invoice rows, check the row count, compare each field and independently calculate quantity times unit price and the grand total in a spreadsheet. Test a missing quantity and a leading-zero item code.
- Deliverable: A verified extraction table with a short exception log.
- Knowledge check: Identify duplicate, missing and incorrectly typed records before approving an import.
07Machine Learning and Model Evaluation
- Learning objective: Distinguish common learning tasks and choose suitable evaluation measures.
- Lesson content: Supervised learning uses labelled examples; unsupervised learning seeks structure in unlabelled data. Classification predicts a category, regression predicts a numerical value, and clustering groups similar examples. Features are model inputs; labels are the target outputs in supervised learning. Training, validation and test data serve different purposes. Data leakage occurs when information unavailable in the real task enters training or evaluation, making performance look better than it is.
- Use accuracy, precision and recall for suitable classification tasks and an error measure such as mean absolute error for regression. Class imbalance can make accuracy misleading. Overfitting means fitting training patterns that do not generalize well. Evaluate the intended use, not only one headline metric.
- Practice: Review a fictional spam classifier with 20 test messages. Count true positives, false positives, true negatives and false negatives. Calculate accuracy, precision and recall, then discuss the cost of blocking a legitimate message.
- Deliverable: An evaluation worksheet and a plain-language recommendation.
- Knowledge check: Explain why a separate test set and a baseline comparison matter.
08Designing Reliable AI Workflows
- Learning objective: Design an AI-assisted workflow with validation and a clear review step.
- Lesson content: Map a workflow as input, preparation, AI task, validation, human approval and output. Use conventional rules for deterministic requirements such as mandatory fields and numerical checks. Use AI where interpretation or generation adds value. Retrieval-augmented generation supplies relevant retrieved material to a model; it can improve grounding but does not guarantee correctness. An agent-style workflow may choose tools or actions, so define its permissions and stop conditions explicitly.
- Treat text in external documents as data, not automatically as trusted instructions. Check for prompt injection, conflicting requests and unintended disclosure. Build fallbacks for missing data, invalid output and unavailable tools. Prototype with fictional data before connecting live systems.
- Practice: Design a support-enquiry workflow that classifies a message and drafts a reply. Route uncertain cases to a person; require approval before sending. Test an empty message, an off-topic message and an embedded instruction to ignore the workflow rules.
- Deliverable: A workflow diagram, input/output schema and exception-handling checklist.
- Knowledge check: Identify which actions need explicit authorization and where the workflow should stop.
09Responsible AI, Privacy and Human Oversight
- Learning objective: Identify misuse and quality risks and assign practical controls.
- Lesson content: Consider privacy, bias, fairness, transparency, security and accountability in the task being performed. Remove or minimize personal and confidential information before using a tool, and confirm that its settings and organizational policies permit the intended use. Do not assume all AI tools handle data in the same way. Review generated statements about people and avoid unsupported conclusions.
- Bias can arise from data, labels, design choices and how a system is used. Test varied scenarios and document limitations. Human oversight should specify who reviews the output, what they check and when escalation is needed. AI-generated medical, legal or financial material requires appropriate qualified review before being used for consequential decisions.
- Practice: Review a fictional AI-assisted hiring summary. Identify irrelevant personal details, unsupported rankings and missing evidence. Redesign the workflow to support a human reviewer with a consistent, job-related checklist.
- Deliverable: A risk register listing scenario, possible harm, control and reviewer.
- Knowledge check: Explain why removing a name alone may not make a record anonymous.
10Capstone: Build and Review an AI-Assisted Solution
- Learning objective: Demonstrate a complete, tested AI-assisted solution to a realistic problem.
- Project brief: Choose a document-summary assistant, course-content assistant or enquiry-classification workflow. Use public or fictional data. Define the audience, problem, success criteria and permitted actions. Create a prompt or workflow, specify the expected output structure and build a small demonstration. Coding is optional.
- Testing: Prepare at least ten test cases including typical inputs, missing information, ambiguous requests and conflicting instructions. Record expected behaviour, actual output, factual errors, format errors and required corrections. Where repeatability matters, run selected cases more than once. Include a manual baseline and note the tool, settings and date used.
- Deliverables: A problem brief, prompt/workflow, sample inputs and outputs, test log, risk checklist and five-minute demonstration. Explain limitations and the review steps required before real use.
- Proposed assessment rubric: Problem definition 15%; workflow and prompt quality 25%; output quality and evidence checks 25%; testing and responsible use 25%; presentation 10%. This rubric is a course-design proposal, not a confirmed certification policy.
- Reflection: Describe one failure, the change you made and the evidence that the change improved the result.
05 / Before joining
Eligibility & requirements
Who can enquire?
- Students, graduates, working professionals, educators, administrative teams, content creators and small-business teams seeking a practical introduction to AI. Suitable for learners without a programming background.
What do you need?
- Basic computer and internet skills.
- ability to read and write in English
- access to a computer, internet connection and spreadsheet application. No previous AI experience or coding knowledge is required. Practical exercises need an AI assistant approved by the training provider; use public or fictional data.
06 / Fee and enrolment
Program fees & admission
Training program fee
Enquire for Price
07 / Need more clarity?
Frequently asked questions
Who is this course for?
This course is designed for students, graduates, educators, professionals and business teams who want a practical introduction to AI. No previous AI experience is required.
Do I need coding experience?
No. The core exercises use prompts, documents, spreadsheets and workflow planning. Coding is optional for the final project.
What will I learn?
You will learn AI fundamentals, generative AI, prompt design, communication and content workflows, data extraction, basic machine-learning evaluation, responsible use and project testing.
How long is the course?
The proposed course contains 30 learning hours across ten modules, including practical exercises and a four-hour capstone module. The training provider will confirm the timetable and access period.
How will the course be delivered?
The course is planned for online delivery. The training provider will confirm the session schedule, platform and whether recordings or self-paced access are included.
What equipment or software do I need?
You need a computer, internet connection, spreadsheet application and an AI assistant approved by the training provider. Confirm any account or subscription costs before enrolment.
What practical work is included?
You will create a prompt library, review meeting notes, plan a content campaign, verify extracted invoice data, evaluate a sample classifier and design an AI-assisted workflow.
What is the final project?
Choose a document-summary assistant, course-content assistant or enquiry-classification workflow. Submit a brief, prompts or workflow, test cases, outputs, risk checks and a short demonstration.
How is learning assessed?
The proposed approach uses module knowledge checks, practical deliverables and a final project. The capstone rubric covers problem definition, design, output quality, testing, responsible use and presentation. Final pass criteria must be confirmed by the provider.
Will I receive a certificate?
Certificate availability and completion requirements have not been confirmed. Please check with the training provider before enrolment.
Can I use my own business information during exercises?
Use public or fictional data for the exercises. Use real business information only when organizational policy, permissions and the selected tool permit it; minimize personal and confidential details.
Does this course qualify me as an AI engineer?
This course develops foundational knowledge and practical AI literacy. It is an introduction rather than a complete AI engineering programme; advanced development requires further study and programming practice.
What are the fee and upcoming dates?
Fees, start dates and enrolment details will be confirmed by the training provider. These details are intentionally left unset in this course draft.