A Teacher’s Classroom Friction Can Become a Small Learning App

45 minutes in which a teacher showed how to make an app

Teachers without programming experience can now use artificial intelligence apps to create other apps for their classes. On August 29, 2026, at the VietTESOL International Conference in Da Nang, Vietnam, Dr. Anthony Pickot introduced this approach for English teachers in a 45-minute, hands-on session in a room for around 100 people.

Pickot described using AI apps to make other apps. He defined it as a way for people with no programming experience to create apps themselves.

His demonstration used Dictaflow, an app designed for integration into intensive listening or reading lessons. The session covered not only its functions, but also the order in which it fits into a lesson and the educational rationale for using it.

Pickot said he had taught the same production method to students in the Master’s programme in English Language Teaching at Manchester Metropolitan University. However, a public trial link for Dictaflow and comparative results on learning outcomes were not disclosed. It is also not confirmed whether it is the same app as products with the same name in app stores.

In Korea, the Ulsan Metropolitan Office of Education, the local education authority, has formally operated the Our Children AI Teaching and Learning Service (우리아이 인공지능 교수학습 서비스), built with participation from working teachers, since March 1, 2025. The service began with 101 lesson models, and its current official page lists around 150 teaching and learning plans. It is a Korean example of teacher-made tools being used in public education after security review and institutional operation.

The repeated work of one English teacher can change first

Consider a teacher responsible for a Grade 8 English class. Each week, to reteach connected speech and short function words that students often miss in listening tests, the teacher searches for videos and audio.

The teacher copies video links, records the start and end of each clip, and moves transcripts into documents. They create gap-fill questions, save answer sheets and student files separately, then make materials again for the following week after reviewing the words students got wrong.

The problem is that this work is not connected to each student’s difficulty. Whether a student could not hear a sound, did not know its spelling, or understood only after seeing subtitles is scattered across paper and multiple files. It cannot be reused in the next lesson.

A small app can bring this flow together. The teacher enters audio and the correct sentence, then divides it into segments. Students proceed through listening without subtitles, dictation, answer checking, and listening again. Instead of names, the app can use temporary numbers and show the teacher only the words that many students missed.

There is no need to build a school-wide learning platform from the start. A first version can have three screens: “add material,” “student completes activity,” and “view common wrong answers.” Features requiring judgement, such as score prediction, automated counselling, or student-record writing, are safer to postpone.

The opportunity became visible in Korea when Ulsan’s service entered formal operation on March 1, 2025.

An app does not remove the teacher’s work. A person still needs to check video difficulty and copyright, automatic-caption errors, and expressions that may be inappropriate for students. Printed activities are also needed for students without devices or with weak connections.

So before deciding what to collect, decide what not to collect. If the purpose can be achieved by storing only whether an answer was correct, without collecting student names or free-form conversation, start there. Teachers must be able to edit AI-generated questions and explanations before they are released.

Other schools started with small problems

US elementary teacher Brian Dang moved lessons that had been handled with paper and spreadsheets into Microsoft Power Apps and created 8-bit Classroom. He built small functions such as mixed-number visualisation and a reward bank for learning, while students also built topic apps in teams.

In a case published by Microsoft, students who used visual materials achieved an answer rate above 80% on mixed-number problems. This was not a study with a comparison group. But it is clearly an example of a teacher turning an explanation problem in one unit into a tool, then extending it into an activity that other teachers could reuse.

At Deerfield Community School in New Hampshire, US, 15 Grade 7 students used the free tool MIT App Inventor to make apps about the water cycle and water pollution. Working in pairs, they built for two weeks and presented their completed work to MIT and other students.

The tool had no usage fee, but the work still required Chromebooks, tablets, and teacher time. After students continued building during lunch and at home, the school decided to run a summer app-building workshop. The case shows that a short class can reach completion and presentation when the topic is narrow.

McMinn County Schools in Tennessee, US, which manages nine schools and around 5,000 students, turned classroom inspections, field-trip approvals, and leave requests into small online forms and apps. Stephen Brady, the supervisor of elementary education and assessment coordinator, set them up so principals and instructional coaches could open materials and feedback forms together on their phones.

In a supplier case study published by Jotform, the district operated around 150 forms and reduced field-trip approval time by an average of 87.5%. The value of its institutional contract was not disclosed. Still, the approach is useful: rather than replacing a large system from the start, it changed the route of one approval task and expanded from there across the school system.

Four things you could build from this

1. A classroom-friction inbox for teachers

This service lets teachers record one repeated classroom frustration in a sentence just after class. It groups similar records and shows problems that may be worth turning into an app.

Its user is a middle-school English teacher who remakes listening materials and worksheets every week.

Now that Korea has an example of working teachers participating in tool creation, a school’s first task may be finding repeated problems rather than buying finished products.

The first screen contains only three fields: “What did I repeat today?”, “How many students experienced it?”, and “Will it happen again in the next lesson?”

2. A three-screen lesson-app design helper

This service takes a teacher’s description of a problem and produces three needed screens and the student’s sequence of use in simple language.

Its user is an English academy instructor or after-school teacher who wants to test a supplementary-class tool before hiring a developer.

Because examples of making small apps by describing functions in words have appeared in teacher training and conference presentations, there is now a need for an intermediate tool that turns an idea into a document a maker can understand.

The first screen asks: “Who is it for?”, “Where do they get stuck?”, and “What one action should the app make them do?”

3. A marketplace for reviewed lesson-activity templates

This service lets teachers share the sequence and blank screens of listening and reading activities they have made. Other teachers can insert their own materials and reuse them.

Its user is a temporary English teacher who does not have time to create an app from scratch but wants to use a reviewed activity from another teacher at the same grade level.

As in the Ulsan case, when lesson plans grow from 101 to around 150 types, review and reuse become more important than adding new functions.

The first screen shows grade level, activity time, required devices, whether student login is needed, and the date of teacher review first.

4. A student app-project progress board

This service helps a teacher see where each student team is, from choosing a problem through screen planning, test use, and presentation.

Its user is a middle-school teacher running a two-week app-building activity in a free-semester programme or science class.

As the Deerfield example shows, completion is possible in a short period. But teachers need to monitor the scope of several teams’ topics and whether they use personal information at the same time.

The first screen shows four statuses for each team—“problem chosen,” “first screen made,” “tested by a classmate,” and “presentation ready”—along with a teacher confirmation button.

Why this matters where you are

You can check whether teachers in your own setting repeatedly remake the same materials, forms, or approval steps. The institutions, review process, available devices, and privacy requirements may differ from those in the examples. Start by identifying one repeatable task that can work without student names or other personal information.

What to check in 30 minutes today

Call one English teacher and ask to see a worksheet or file they have remade at least three times in the past month. If they can describe the input material and finished result in one sentence each, and the task can be handled without student names or personal information, it may be a problem worth testing with a small app.