Büşra Aydın Erkuş

Term project for voice recognition and assistant training

I evaluated the quality of the responses provided by the AI models based on multidimensional metrics.
I participated in Siri projects for Apple, testing linguistic and experiential consistency. (For both English and Turkish models)
I contributed to the training and development of the models with feedback.
From a software testing perspective, I reported on how the models performed in edge-case scenarios.

Participated in multilingual LLM tasks including text evaluation, transcription, and app testing.

Annotated and labeled large datasets for machine learning pipelines.

Evaluated social media ads and app UIs for relevance and user experience.

Voice data collection for voice recognition and assistant training.