September places a particular kind of pressure on educators. There are new learners to understand, specifications to revisit, schemes of work to prepare and first lessons that need to feel purposeful from the start.

For many teachers, the challenge is not knowing how to teach. It is finding the time to turn curriculum knowledge, learner context and assessment requirements into well-sequenced plans before teaching begins.

That was the focus of the Deep Dive session Start Strong with TeacherMatic. Rather than presenting AI as a collection of disconnected tools, the session followed a realistic planning journey:

1. Scheme of Work / Curriculum Plan
2. Student Profile
3. Lesson Plan
4. Classroom Questions
5. Learning Activities
6. Review and refine

The strongest message was simple: TeacherMatic is most useful when it supports an educator’s existing workflow. It can remove the blank page, speed up first drafting and offer structured prompts, but it does not remove the need for teacher judgement.

Start With The Curriculum Roadmap

The session began with the Scheme of Work/Curriculum Plan Generator, using Pearson BTEC Level 3 Unit 3: Using Social Media in Business as the running example.

This was a useful choice because it showed how one unit could be carried through the whole planning process. The Scheme of Work established the curriculum route. The Student Profile then shaped how the first session might be approached. The Lesson Plan turned the opening point in that route into a teachable session. Classroom Questions and Learning Activities helped bring the lesson to life.

A strong Scheme of Work is not just a table of topics. It sets out how learning should develop, where assessment belongs, what needs to come first and how each session contributes to the wider curriculum journey.

TeacherMatic can accelerate that first draft by working from a specification, a document, an existing plan or a topic list. But the educator still needs to review the sequence, check assessment logic, consider local timetable realities and make sure the result fits the learners and setting.

Practical idea

If you are working from a large awarding organisation document, extract only the relevant unit pages before using them as source material. Focused input usually produces a more useful first draft and keeps the planning centred on what you actually need to teach.

Plan For The Learners, Not Just The Unit

A Scheme of Work tells us what needs to be taught. It does not tell us enough about who is in the room.

That is why the next stage of the workflow used the newly updated Student Profile Generator. At the start of the year, educators may only have partial information: induction notes, initial assessment, support needs, prior attainment or early observations. Even so, that information can shape better planning.

In the Deep Dive example, the learner was confident and talkative, familiar with Instagram and TikTok through personal use, but less confident with formal written work. That immediately changes the first lesson design. Instead of beginning with extended written analysis, it makes sense to start with discussion, recognisable examples and collaborative activity, then build towards the written assessment expectations.

This is adaptive teaching in practical terms: not creating a separate curriculum, but thinking carefully about how learners can access the same intended learning.

The screenshot also shows an important responsible-AI point. The generator prompts users to avoid personally identifiable information. That matters. AI-supported planning should use enough context to improve teaching decisions without exposing unnecessary learner data.

Turn The Roadmap Into A First Lesson

The Lesson Plan Generator then moved the workflow from curriculum planning into classroom planning.

The first lesson has a particular job to do. It introduces the topic, but it also sets expectations, starts relationships and gives learners an early sense of whether the unit feels relevant. For Unit 3, the session did not begin with formal report writing. It began with what learners already understood: their own use of social media.

Using Greggs as a recognisable business example gave the lesson a concrete starting point. Learners could compare personal social media habits with how a business uses tone, timing, audience awareness and brand identity strategically.

The pedagogical value here is not simply that AI can write a lesson plan. It is that the right context helps produce a more usable plan: topic, sector, lesson length, learner profile, environment and the kind of teaching approach needed for that group.

A lesson plan should not sit in isolation. It should make sense within the wider sequence of learning and the learners’ starting points.

Prepare The Questions Before The Lesson Begins

Questioning is one of the simplest parts of teaching to underestimate. In a live classroom, educators are explaining, listening, watching the room, adjusting pace and checking understanding. Planning a few strong questions in advance can make a noticeable difference.

For the first session, open questions made more sense than closed recall questions. Learners had not yet been taught the formal content, so the purpose was to surface existing experience and begin shifting thinking from personal use to business strategy.

Later in the unit, the same generator could be used differently: closed questions for retrieval practice, quick checks for understanding or a short starter quiz. The key point is that the educator decides the pedagogical purpose first, then uses the generator to support that purpose.

Use AI To Get Out Of Teaching Autopilot

The Learning Activities Generator raised one of the most useful points in the Deep Dive: sometimes the challenge is not teaching unfamiliar content. It is teaching familiar content in a way that still feels fresh and purposeful.

Experienced educators rightly build up examples, routines and activities that work. But when a unit has been taught many times, it is easy to default to the same route through the same material. AI can be useful here as creative provocation.

For the social media session, the prompt pack suggested activities such as a “guess the brand” challenge using the tone of social posts, a quick sort between personal habit and business strategy, or a “you’re hired” scenario where learners act as junior social media assistants for a local business.

You do not need to use every idea. A better approach is often to ask for more activities than you need, then select and adapt the strongest ones. That keeps the professional decision with the educator while still widening the range of options.

Review, Refine And Keep The Human In The Loop

The Deep Dive also demonstrated why review matters. A generated output should never be treated as finished just because it looks complete. It needs checking for accuracy, level, tone, assessment fit and suitability for the learners.

This is where the 80/20 idea is useful. TeacherMatic can do a substantial amount of first-draft work, but the remaining educator input is not a small afterthought. It is the quality assurance stage. It is where professional knowledge is applied.

During the session, the Refine function was used when the first Scheme of Work output needed improvement. That moment is important because it models responsible AI use: generate, review, challenge and refine. The educator remains accountable for the final resource.

What This Means For Educators

The strongest use of TeacherMatic is not simply generating individual resources faster. It is using AI across a connected teaching workflow while keeping professional judgement at every stage.

• Start with the curriculum. Build or review the Scheme of Work before jumping into individual lesson resources.
• Use focused source material. Provide the relevant unit, topic or section rather than unnecessary documents.
• Add learner context where it helps. Early knowledge of the group can change the suitability of questions, activities and lesson structure.
• Generate options, not final answers. Ask for several questions or activities, then choose and adapt what fits your class.
• Do not skip refinement. Every generated result should be checked, challenged and made fit for purpose.

Start Strong Without Starting From A Blank Page

The goal is not to automate teaching. It is to reduce the time educators lose to first-draft admin so that more time can be spent on the decisions that matter.

A Scheme of Work still needs curriculum expertise. A Student Profile still requires thoughtful handling of learner information. A Lesson Plan still needs to work in a real classroom. A question or activity still needs an educator to decide whether it will help this group learn.

That is where TeacherMatic is most valuable: not as a replacement for professional judgement, but as a way to help educators start stronger, move faster and keep their expertise focused where it has the greatest impact.