Sofya Durneva

Audits and rates AI-generated text and speech, designs rubrics/guidelines, and performs cross-lingual error analysis to support linguistic quality improvements. Provides translation, proofreading, editing, and language instruction in English and Russian.

Collected and processed data via hotline and chat, devised a semi-automated markup pipeline, and translated materials into English. Conducted open-source research and surveyed municipal deputies by phone.

Designed and implemented language quality assessment protocols for multilingual AI voice-to-voice interpretation products, evaluating speech recognition, machine translation, and text-to-speech quality. Presented prioritized linguistic error insights to ML engineering and business teams and performed semi-automated data collection for training.

Worked on the Database of Semantic Shifts (datsemshift.ru) by verifying and adding semantic shifts and representations. Annotated data for an ML model and parsed dictionaries using Python.

Developed conversational design for an NLP-based bot on a bot-builder platform, debugging and testing the bots. Produced analytical reports and recommendations for further bot development.

Prepared and conducted classes, exams, and course plans for linguistics and academic English disciplines. Supported teaching across multiple course formats including Introduction to Linguistics and academic writing courses.

Supported early development of markup for an email dataset intended for training a neural network. Helped prepare structured labeled data for model training.

Adapted a transliteration web app for Dargin languages, contributed new texts to the Kubachi language corpus, and participated in fieldwork in Dagestan. Supported corpus building and language data preparation tasks.

Prepared tax refund documents for foreign clients, explained the refund process, and answered client questions. Maintained records in MS Excel and communicated by email and phone, mainly in English and Chinese.

Master’s program in Linguistic Theory and Language Description. Coursework completed and thesis remained pending as of the academic leave since 2022.

Bachelor’s degree in Theoretical and Computational Linguistics at the National Research University “Higher School of Economics” in Moscow.