A Step by Step Guide for Creating Courses with AI
For teachers, trainers, and course designers who want to create useful courses faster and with more confidence.
Introduction
AI can be more than just a tool for quick answers. It can be your partner in building a full course. In this guide, I will show you a simple, step‑by‑step process for how to use AI in course creation. To make it easier to follow, we will use one example: adult immigrant students at a B1–B2 level who are preparing to work in healthcare support roles in an English‑speaking country. These students need English for practical communication at work and in daily life. As we go through the steps, you will see how a teacher can ask AI for exactly what they need and how to check the quality of the output.
Step 1: Brainstorm real‑world language tasks
The first step is to identify what your students really need to do in English. AI can give you a list of real‑world tasks. For healthcare support workers, this might include:
Greeting a patient politely
Taking basic medical information
Explaining hospital procedures
Writing short notes for a supervisor
Understanding instructions from a nurse or doctor
Prompt example:
“You are an instructional designer. Brainstorm a list of 25–30 real‑world tasks for B1–B2 adult immigrants training as healthcare support staff. Group them by simple, medium, and difficult. For each task, explain the purpose, the context, and the expected output.”
The result should be a practical task list that can guide your whole course.
Step 2: Match tasks with learning outcomes
Next, you need to check which official learning outcomes match your list of tasks. Many teachers use Pearson’s Global Scale of English (GSE) or the CEFR. For example, the task “Take basic medical information” might connect to outcomes like “Can ask questions about someone’s personal details” or “Can understand simple factual information in short texts.”
Prompt example:
“You are a curriculum assistant. Cross‑reference this list of tasks with GSE outcomes at the B1–B2 level. For each task, show 3–5 outcomes it supports in a table. Include a column for evidence of how the task meets the outcome.”
You can ask AI to create a table that shows each task and the outcomes it supports. This makes your course measurable and connected to international standards.
Step 3: Build a scope and sequence
Now you need to decide the order of units. AI can suggest a logical scope and sequence. For example, you might begin with polite greetings and introductions, then move to gathering information, and later to explaining more complex procedures. Each unit builds on the last one.
Prompt example:
“You are a course architect. Create a 6‑unit course for B1–B2 healthcare support workers. Each unit should have 6 lessons. For each lesson, list the main task, outcomes, and how it connects to the next lesson.”
The AI should produce a clear table with units, lessons, outcomes, and main tasks.
Step 4: Analyze authentic texts and find useful language
Authentic model texts are real examples, such as a hospital intake form, a short workplace email, or a sample dialogue between a nurse and patient. AI can help you analyze these texts for useful vocabulary and expressions. For example:
Vocabulary: “blood pressure, appointment, symptom, prescription”
Expressions: “Can you please tell me…?” or “Let me check that for you.”
Prompt example:
“You are a language analyst. From these sample texts (pasted below), identify 12–15 key vocabulary items and 8–10 functional expressions that healthcare support workers need. For each, provide meaning, IPA pronunciation, part of speech, and 4 CCQs (yes/no, no, either/or, and open). Include notes on possible L1 interference.”
AI can also give you a detailed language analysis, including meaning, pronunciation, and possible learner difficulties. For example, many learners might confuse “prescription” with “description.” Including these notes helps prevent mistakes.
Step 5: Create leveled practice texts
After collecting authentic examples, AI can help you write leveled texts for reading and listening. These texts should have the same purpose as the authentic models but be adjusted to student level. For example:
Easy text: A short patient dialogue with simple vocabulary.
Core text: A longer dialogue where a nurse gives basic instructions.
Challenging text: A detailed hospital procedure explained in plain language.
Prompt example:
“You are a materials writer. Create three leveled practice texts (easy, core, and challenging) for B1–B2 learners based on this authentic hospital dialogue. Include a short vocabulary pre‑teach list, a gist question, 3 detail questions, and one personalization question.”
You can also ask AI to provide comprehension questions and short personalization tasks, such as: “Have you ever filled out a form like this in your own country?”
Step 6: Produce lesson plans and materials
Finally, you can ask AI to create lesson plans based on the scope and sequence. For example: “Write a 60‑minute lesson plan for B1–B2 healthcare support learners. The lesson aim is to take a patient’s basic medical information. Include warm‑up, vocabulary, practice, role‑play, and feedback stages.”
Prompt example:
“You are a lesson designer. Write a detailed 60‑minute lesson plan for B1–B2 healthcare support learners. The lesson aim is to practice taking a patient’s medical history. Include: stages, timings, teacher instructions, ICQs, differentiation for faster/slower learners, and anticipated problems with solutions. Provide student worksheets with role‑play cards.”
AI should give you a full plan with timings, clear teacher instructions, and materials like role‑play cards or short dialogues. You can then adapt these materials to fit your style and your students.
Conclusion
By following these six steps, you can move from big ideas to concrete lesson plans with the help of AI. The process is simple:
Brainstorm tasks.
Match them to outcomes.
Create a logical sequence.
Analyze authentic language.
Produce leveled texts.
Design lessons and materials.
The key is to give AI clear information about your learners and to always check the quality of the output. Used this way, AI is not replacing teachers but supporting them—helping you create courses that are practical, relevant, and ready for the classroom.