olesel@icloud.com

Kocherov Kirill · CPO
Larichev Pavel · PO
Langueva Ksenia · Art-director
Selivanov Oleg · Senior product designer
Ermochkina Tamara · UX-writer
Grigorenko Andrey · Motion-designer
and others

A mobile app that helps teachers analyze their lessons and develop their professional skills using artificial intelligence

jul 2024–nov 2025

Senior Product Designer

edTech · mobile app

AI teaching assistant · SberEducation

Context

AI Teacher Assistant is a product that helps teachers analyze their lessons and develop their professional skills using artificial intelligence.
The service analyzes an audio recording of the lesson and generates a report with metrics and recommendations: lesson outline, distribution of speaking time between the teacher and students, speaking speed, emotional tone, frequently used words, and the use of pedagogical and sociological techniques.
Before the mobile app was released, the product was available only as a web version. To get a report, teachers had to record the lesson on their phone, transfer the audio file to a computer, and upload it to the web service.
The mobile app was designed to remove these barriers and make using the service a natural part of a teacher’s daily workflow

Problem

Using the service required a lengthy and cumbersome process involving multiple devices. To obtain a report, the teacher had to:

1. pick up their phone
2. open the voice recorder
3. start recording
4. teach the lesson
5. stop recording
6. find a way to transfer the file to a computer
7. open the web service
8. log in
9. upload the audio file
10. submit the recording for analysis

After the recording was processed, the report was only available in the web version.
This process required extra effort and time, so at every stage, users could lose motivation to continue using the product.
The service’s core value—lesson analytics and recommendations for teacher development—was hidden behind a complex technical workflow

The problem wasn't with the product analytics. The main hurdle lay between recording the lesson and uploading the file. The more steps required to upload the recording, the more likely it was that the teacher would postpone analyzing the lesson or skip it altogether

The goal of the mobile app is to streamline the process from recording a lesson to receiving a report. That’s why the app was designed around a single key action: recording a lesson and sending it for analysis

Solution

Since the launch of the mobile app, the process has become much simpler.
The instructor opens the app, taps the record button, stops recording after the lesson, and sends the file for analysis. When the report is ready, they receive a push notification.
The process of obtaining analytics has been streamlined into a single, seamless workflow within a single device

Restrictions

AI lesson analysis was already available in the web product, so the mobile solution had to integrate into the existing audio processing logic rather than create a new process. At the same time, the primary user scenario took place in an offline lesson context, where the teacher cannot afford to spend time on complex actions or settings. This required designing the shortest possible path to recording and submitting a lesson for analysis, without increasing the cognitive load on the user

Result

After the mobile app was launched, the product began to be used in a different way.
The AI assistant became a teacher’s pocket tool.
The product’s core features—recording lessons and generating analytics—became accessible in a single, simple workflow without the need to switch between devices

Conclusion

Sometimes the key design challenge isn’t to improve the interface, but to restructure the user experience around the core task.
In this project, the product’s value became clear once the path between the user’s action and the result was streamlined into a single, simple scenario

thanks!