Computational Linguist and Japanese NLP Annotator

Hiromi Oyama

Present: *** ******, *******-****, ********,861-2242, Japan,

Tel: 090-****-**** E-mail: ******.*****@*****.***. Profile:

Computational Linguist / Analytical Linguist with a Ph.D. in Computational Linguistics and extensive experience at leading tech companies (Google, Apple, LINE). Leverages deep expertise in Japanese linguistics, ASR/NLP data curation, and Python-based data analysis to build and refine language AI models (LLMs, Assistant, ASR). Skilled in taxonomy design, evaluation, error analysis, and cross-functional collaboration between engineering and linguistic teams.

Qualifications:

● Technical Proficiency: Skilled in Python and Ruby, databases as MySQL, statistical computing applications such as R, and tools with over 10 years of experience. Proficient in data manipulation and analysis using pandas, numpy, and other relevant libraries.

● Project Management: Managed vendor teams for data annotation to ensure the efficiency, accuracy, and quality of linguistic data and systems with cross-functional teams for the successful delivery of complex IT projects.

● Problem-Solving: Demonstrated ability to troubleshoot complex technical issues and provide effective solutions, by using text processing scripts to reduce human evaluation time.

● Certifications: Python 3 Engineer Certified Data Analysis (d4yr-2wb8-s3fh)

(Feb.2024)

● Language Proficiency: TOEIC 905 (Dec. 2020),EIKEN Pre-1, French (Beginner) Education:

● Nara Institute of Science and Technology (NAIST), Japan Ph.D in Computational Linguistics (Information Science)

● Lancaster University, U.K.

MA in Language Teaching in Applied Linguistics

● The Prefectural University of Kumamoto (PUK), Japan MA in Japanese Linguistics and Teaching Japanese as a foreign language

● PUK, Japan

BA in English Linguistics and Literature

Work Experience:

Apr.2026 -- Present Coto College Japanese language teacher& Academic management

● Provide Japanese language instruction while overseeing academic operations, faculty management, curriculum development, and institutional initiatives.

● Lead the preparation of institutional documentation required for accreditation as a Certified Japanese Language Education Institution under MEXT, ensuring full compliance with national standards.

● Prototyped an AI-powered Japanese essay correction application utilizing LLM prompt engineering, enabling automated grammatical error analysis and natural phrasing suggestions tailored to learners' proficiency levels.

● Integrate LLM tools and generative AI into curriculum design to streamline custom material generation and personalized feedback mechanisms.

● Implement data-driven academic planning, assessment tracking, and administrative workflow automation to optimize operational efficiency.

● Design and deliver faculty training programs focusing on advanced pedagogy, classroom management, and the integration of EdTech/AI in language instruction. Outcome:

● Advanced institutional readiness for MEXT accreditation by developing compliant curricula, operational policies, and standardized documentation.

● Demonstrated practical LLM integration in language education by developing an AI essay correction prototype, enhancing feedback speed and instructional accuracy.

● Increased operational efficiency across academic and administrative teams through process standardization and technology integration. Nov.2024 – Mar.2026 NiCE Japan Co., Ltd. Solution Consultant, Analytics

● Designed and configured analytics solutions tailored for call centers, enabling sentiment analysis and root cause investigation from multi-channel data.

● Delivered training and ongoing support to ensure effective adoption.

● Built a cloud-based demo web application (Python + Flask) combining third-party APIs to automate speech recognition, sentiment analysis, and visualization for customer proposals.

Outcome:

-Supported client evaluation of analytics solutions by demonstrating how customer sentiment and operational issues can be identified from call center data. Oct.2023 -- Oct.2024 Aceolution Japan (dispatched to Google Japan) Language Specialist

● Evaluated Gemini generative AI outputs for accuracy, factuality, and clarity as part of a structured quality assurance process.

● Curated and validated datasets to improve ASR and NLP performance in Google Assistant.

● Conducted QA and bug-fixing for Japanese corpus data.

● Collaborated across teams to ensure cross-lingual integration.

● Delivered evaluations within defined schedules while maintaining consistency and quality standards.

Outcome:

-Supported efforts to improve the reliability and evaluation quality of generative AI systems used in production environments.

Feb.2022 -- Sep.2023Uniphore Technologies Japan Business Analyst

● Configured ASR and NLP engines for enterprise solutions.

● Designed conversation processes and analyzed call transcripts to extract patterns and summary entities.

● Conducted root cause analysis and QA to improve recognition accuracy and overall performance.

● Gathered client requirements and provided recommendations for product design, feature prioritization and UX improvements

● Worked closely with engineering and product teams to validate requirements and support solution delivery.

Outcome:

-Contributed to improved ASR/NLP performance and more accurate conversation insights, helping clients evaluate the solution for future enterprise deployments. Apr.2021 -- Feb.2022 PeopleCert International Ltd. Business Development Assistant

● Conducted market research to identify new partners in IT and language certification domains within the Japan market.

● Identified and approached new partner candidates, supporting initial outreach and qualification.

● Facilitated onboarding processes and assisted with ongoing partner relationship management.

● Supported sales and business development managers with market insights, partner coordination, and operational tasks to expand the Japan market. Outcome:

-Supported the development of a partner ecosystem in Japan by contributing market research and partner onboarding efforts.

Jul.2020 -- Jan.2021 LINE Japan Linguist

● Built and annotated Japanese speech corpora for ASR model training used in Clova smart speaker.

● Performed prosodic annotation, including pitch and intonation labelling, to enrich speech data and support recognition quality analysis.

● Carried out linguistic quality testing and ASR error analysis to identify systematic recognition issues specific to Japanese.

● Collaborated closely with ASR engineers and product teams to improve data quality, annotation guidelines, and overall ASR performance. Outcome:

-Supported improvements in accuracy and robustness of Japanese ASR for Clova by strengthening training data quality, lexical coverage, and linguistic evaluation processes. Mar.2018 -- Mar.2019 Google Japan (Temp via Robert Walters) Machine Intelligence Analytical Linguistic Project Manager

● Curated and validated large-scale Japanese datasets to optimize NLP performance for Google Assistant.

● Designed and evaluated Japanese user queries for QA systems, assessing intent coverage and response accuracy.

● Conducted corpus-level QA, identified data issues. Outcome:

-Supported improvements in Japanese NLP quality for Japanese-language Assistant features by strengthening data quality, evaluation processes and cross-team coordination. Apr. 2016 -- Mar. 2018 Apple Japan Linguistic Expert

● Annotated Japanese input system data to improve usability and model performance.

● Developed detailed annotation guidelines and coordinate task execution with external vendors.

● Conducted quality assurance audits to ensure annotation consistency, accuracy, and overall data integrity.

Outcome:

-Delivered high-quality, well-documented annotated datasets and standardized guidelines, enabling stable downstream system improvement and more consistent input behavior. Technical skills:

Development Environments: Windows, MacOS, Unix, Docker, Conda, Git

Programming language: Ruby, Python, Shell script

Database: PostgreSQL, SQLite

Statistical computing application: R

Tools: Sklearn, Pandas, Numpy, Regex

Applications: API integration, web app development (Flask), Excel (advanced), PowerPoint

Projects and contributions:

Developed Web application to search for Kumamoto dialect sentences

Developed a cloud-based Web application using Flask and Python, integrating APIs for automated speech recognition and sentiment analysis.

Organized "Study support project for language learners of Japanese for ubiquitous computing society". Participated in Nara Institute of Science and Technology (NAIST) Creative and International Competitiveness Project (CICP2008)(2008.11-2009.3) Portfolio:

Notion:

https://nonstop-gibbon-7df.notion.site/My-tech-portfolio-26ef18352eb380b9bf29c145c 2697e6b

Language:

Japanese -- Native

English -- Fluent

Credentials:

Passed the Japanese Language Teaching Competency Test authorized by the Japanese Ministry of Education

TOEIC online IP test score 905 (Dec. 2020)

Grade Pre-1st in Test in Practical English Proficiency (STEP)

Contact this candidate