08/17/2026 updated
TN
100 % available
AI Training & Evaluation Specialist | LLM Data Quality Expert
Dong Thap, Vietnam
Only remote
Graphic Design Program - FPT Arena Multimedia (2021-2023)LLM Response Evaluation & RLHF
Expertise in evaluating large language model outputs using multi-dimensional rubrics covering factual accuracy, completeness, reasoning quality, tone, safety, and instruction compliance. Proven track record in RLHF preference annotation and comparative response ranking across globally distributed AI data platforms.
AI Data Annotation & Quality Assurance
Extensive experience in classification, validation, and QA workflows for multilingual AI training pipelines. Application of annotation taxonomies and project rubrics to produce clean, consistent training data for downstream model fine-tuning, including multimodal and structured output tasks.
Prompt Engineering & Instruction Tuning
Hands-on experience designing and evaluating instruction-tuning datasets and structured outputs for LLM training workflows. Contribution to OpenAI-focused instruction-tuning projects, including generating and reviewing AI-generated Excel deliverables for format consistency and instruction compliance.
Hallucination Detection & Safety Alignment
Identification and documentation of failure patterns including hallucinations, constraint violations, logical gaps, incomplete responses, and unsafe or harmful phrasing in AI-generated content.
Vietnamese-English Localization QA
Delivery of Vietnamese-English translation and localization review focusing on fluency, semantic accuracy, terminology consistency, and cultural appropriateness for AI data platforms and language technology companies.
Search Quality Evaluation
Rating of search engine results using detailed guidelines covering query intent classification, result relevance, localization fit, and user satisfaction signals. Evaluation of landing pages for trustworthiness and E-A-T signals.
Audio & Image Quality Review
Assessment of model-generated responses for audio quality tasks, rating output accuracy and labeling consistency. Review of ad creative and image quality signals, identifying low-quality, misleading, or policy-sensitive content using structured rubrics.
Annotation Tool Proficiency
Proficiency with a wide range of annotation and evaluation tools including Label Studio (HumanSignal), ECO, Feather, OneForma Annotation Tool, TELUS/Appen/RWS platforms, Doccano, Labelbox, and Scale AI-style interfaces.
Graphic Design & Visual Content Creation
Creation of social media visuals, print and digital communication materials, educational visuals, and branded content using tools such as Canva, Adobe Photoshop, Adobe Illustrator, Adobe InDesign, and Figma.
Data Pipeline QA & Structured Output Validation
Quality assurance of data pipelines and structured outputs including Excel and spreadsheet-format responses, identifying formatting errors, logical inconsistencies, and usability gaps with actionable feedback for model improvement.
Expertise in evaluating large language model outputs using multi-dimensional rubrics covering factual accuracy, completeness, reasoning quality, tone, safety, and instruction compliance. Proven track record in RLHF preference annotation and comparative response ranking across globally distributed AI data platforms.
AI Data Annotation & Quality Assurance
Extensive experience in classification, validation, and QA workflows for multilingual AI training pipelines. Application of annotation taxonomies and project rubrics to produce clean, consistent training data for downstream model fine-tuning, including multimodal and structured output tasks.
Prompt Engineering & Instruction Tuning
Hands-on experience designing and evaluating instruction-tuning datasets and structured outputs for LLM training workflows. Contribution to OpenAI-focused instruction-tuning projects, including generating and reviewing AI-generated Excel deliverables for format consistency and instruction compliance.
Hallucination Detection & Safety Alignment
Identification and documentation of failure patterns including hallucinations, constraint violations, logical gaps, incomplete responses, and unsafe or harmful phrasing in AI-generated content.
Vietnamese-English Localization QA
Delivery of Vietnamese-English translation and localization review focusing on fluency, semantic accuracy, terminology consistency, and cultural appropriateness for AI data platforms and language technology companies.
Search Quality Evaluation
Rating of search engine results using detailed guidelines covering query intent classification, result relevance, localization fit, and user satisfaction signals. Evaluation of landing pages for trustworthiness and E-A-T signals.
Audio & Image Quality Review
Assessment of model-generated responses for audio quality tasks, rating output accuracy and labeling consistency. Review of ad creative and image quality signals, identifying low-quality, misleading, or policy-sensitive content using structured rubrics.
Annotation Tool Proficiency
Proficiency with a wide range of annotation and evaluation tools including Label Studio (HumanSignal), ECO, Feather, OneForma Annotation Tool, TELUS/Appen/RWS platforms, Doccano, Labelbox, and Scale AI-style interfaces.
Graphic Design & Visual Content Creation
Creation of social media visuals, print and digital communication materials, educational visuals, and branded content using tools such as Canva, Adobe Photoshop, Adobe Illustrator, Adobe InDesign, and Figma.
Data Pipeline QA & Structured Output Validation
Quality assurance of data pipelines and structured outputs including Excel and spreadsheet-format responses, identifying formatting errors, logical inconsistencies, and usability gaps with actionable feedback for model improvement.
Languages
EnglishFluent
Project history
Designed social media and website visuals aligned with brand identity, supported marketing campaigns, KOC/KOL collaborations, and partnership materials, and contributed to community-focused campaigns and charity initiatives.
Delivering LLM training data, RLHF preference annotation, evaluation, and localization QA for AI data platforms and language technology companies worldwide.
Assessed model-generated responses for audio quality tasks, evaluated geographic and map-related AI outputs for spatial relevance and factual correctness, reviewed ad creative and image quality signals, and maintained high inter-rater reliability by following project-specific workflows.