Ade R. - AI Trainer – Coding Expertise & Japanese/Multilingual

Go to Website

Chiba, Japan

Experience

Dec 2024 - Present

11 months

AI Trainer – Coding Expertise & Japanese/Multilingual

Outlier

Designed high-quality coding prompts and model responses that consistently outperformed state-of-the-art outputs from leading large language models.
Conducted expert evaluations of coding prompts and responses in Japanese and English, focusing on Java, Python, C++, and JavaScript, ensuring technical accuracy.
Reviewed team submissions, providing constructive feedback to maintain technical correctness, linguistic quality, and alignment with prompt objectives.
Assessed voice chatbot interactions across Japanese monolingual, English–Japanese bilingual, and Japanese–Korean multilingual scenarios, enhancing overall user experience.
Delivered rubric-aligned evaluations with strong emphasis on consistency, linguistic fluency, and task-specific relevance.
Revised AI-generated Japanese outputs for fluency, clarity, and naturalness, meeting native-speaker expectations.

Apr 2019 - Present

6 years 7 months

Chiba, Japan

Developer at Forecast Center

Weathernews Inc.

Maintained Weathernews’ No.1 forecast accuracy in Japan (Tokyo Shoko Research, 3 consecutive years) by developing an internal monitoring tool to visualize hit rates and detect forecast accuracy declines early.
Developed global forecast generation systems using an AWS technology stack (including S3, Lambda, EC2, CodePipeline) and scalable data pipelines, processing millions of meteorological data points daily.
Built and optimized data workflows using Python for forecast data generation, accuracy calculations, and Lambda integration, ensuring real-time updates for forecasts and reporting.
Automated and enhanced operational workflows by developing a real-time system to generate forecast accuracy reports for ~80–90 tropical cyclones per year and other meteorological events, reducing manual workload by up to 60% and improving reporting efficiency.
Developed front-end dashboards with HTML, CSS, JavaScript, implemented high-speed data processing in C++, and maintained legacy pipelines using Perl, demonstrating adaptability across multiple technology stacks.
Facilitated multilingual reporting and coordination with international teams in Japanese and English, ensuring technical specifications and operational plans were clearly communicated.

May 2017 - Aug 2017

4 months

Bandung, Indonesia

Software Engineering Intern

PT. Dwi Cermati Indonesia

Developed continuous integration (CI) scripts for microservices using Shippable, improving deployment reliability.
Boosted workflow efficiency by ~30% by creating an auto-deployer for the staging server using Ansible and Amazon SQS.
Built unit and functional testing frameworks with Mocha, Chai, and Sinon JS, increasing test coverage and reducing bugs.
Designed end-to-end testing processes using Nightwatch JS and Selenium, ensuring product quality across multiple platforms.
Contributed to the development of a core health insurance product, collaborating with a cross-functional team to deliver key features on time.

Summary

Experienced in full-stack software development and AI model training, I specialize in bridging the gap between language and technology to deliver innovative, high-impact projects.