Introduction Hackathon for ITMO Masters - AI Talent Hub 2025-2027
git clone https://github.com/Alex777Russia/sliding-knowledge-diagnostics.git
cd sliding-knowledge-diagnosticsTo run a contnainer you will need to create .env file in current dir first. Write those variables:
FOLDER_ID=<your_yandex_api_folder_id>
YANDEX_API_KEY=<your_yandex_api_key>
OPENAI_API_KEY=<your_openai_api_key>
Also you need a database in current dir with name data.csv format.
Our data base can be found here.
Start app:
bash docker/build.sh
bash docker/run.shLocal app can be accessed by link http://0.0.0.0:7860 (localhost:7860).
- Voice recording: Students can answer questions by speaking through a microphone
- Audio upload: Ability to upload audio files for transcription
- OpenAI Whisper: Use of the OpenAI Whisper API for high-quality transcription
- Automatic recognition: Audio is automatically transcribed into text
- Flexibility: Supports both text and voice responses
- Russian language: Optimized for recognizing Russian speech
After exam report is generated as PDF file with some plots.
If student's answer is not clear, clarifying question will be generated.




