Kwangsik S.
AI Tutor / Data Annotation Specialist
Experience
AI Tutor / Data Annotation Specialist
Volga Partners
- Delivered advanced AI tutoring and guidance, supporting model reasoning and alignment tasks.
- Contributed to AI’s advanced data annotation project, focusing on Reinforcement Learning from Human Feedback (RLHF).
- Performed detailed annotations and evaluations to improve AI safety, user alignment, and instruction-following capabilities.
- Developed structured feedback loops for fine-tuning large language models in collaboration with researchers and project managers.
Data Annotator
iMerit
- Performed quality control and labeling tasks for AI conversational system data.
- Identified and classified personal information by sensitivity levels.
- Assessed risk levels of text data and assigned appropriate ratings.
- Ensured consistent labeling aligned with annotation guidelines.
AI Adversarial Tester / Prompt Writer
TaskUs
- Designed adversarial prompts to uncover weaknesses, safety risks, and edge cases in LLM behavior.
- Wrote clear, testable, and structured instructions to evaluate AI systems’ performance.
- Collaborated with remote teams to review outputs and escalate flagged issues.
- Generated diverse prompt types across multiple domains (ethical dilemmas, policy, humor, reasoning).
AI Data Trainer
Outlier
- Created high-quality training data for software engineering tasks in Korean.
- Evaluated AI responses and tested for accuracy and bias.
Senior AI Research Scientist
PharmCADD
- Led deep learning projects for protein structure prediction and de novo drug design.
- Applied CNN models, active learning, and RNA structure prediction techniques.
Annotation Analyst (Web Search Evaluator)
Appen
- Verified business and map data accuracy for AI training.
- Supported self-driving car projects with high-quality labeled data.
Linguistic Specialist & Annotation Analyst
Lionbridge
- Conducted software localization, SEO copywriting, NLP linguistic testing, and training data generation for major ML projects.
AI Prompt Engineer
Welocalize & Lionbridge
- Designed, localized, adapted, and validated AI prompts for multiple applications and languages.
- Ensured content safety, cultural relevance, and alignment with project guidelines.
Summary
AI and Machine Learning Specialist with over 10 years of expertise in prompt engineering, adversarial testing, and red-teaming for AI safety and alignment. Demonstrated success in designing high-quality, culturally adapted training data and optimizing AI model performance for global NLP projects. Proficient in Reinforcement Learning from Human Feedback (RLHF), advanced data annotation, and model evaluation, enhancing instruction-following, alignment, and safety of large language models. Adept at collaborating with cross-functional teams to deliver impactful solutions that strengthen AI reliability, user trust, and global applicability.
Languages
Education
Pukyong National University
LL.B. · Law · Busan, Korea, Republic of
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.