Joey D.

Key Skills

Software

Other

Label Studio

Top Subject Matter

Cybersecurity and AI safety (LLM alignment)

LLM alignment and AI training

Top Data Types

Text

Top Task Types

Prompt + Response Writing (SFT)

RLHF

Freelancer Overview

Technical Writer & AI Training Specialist (AI Red-Teaming; SFT/RLHF/DPO-alignment workflows). Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other and Label Studio. Education includes Bachelor degree of Quality Management, University of Ho Chi Minh City (2017) and Introduction to Cybersecurity Certificate, CISCO (2013). AI-training focus includes data types such as Text and labeling workflows including Prompt + Response Writing (SFT) and RLHF.

Labeling Experience

Technical Writer & AI Training Specialist (AI Red-Teaming; SFT/RLHF/DPO-alignment workflows)

OtherTextPrompt + Response Writing (SFT)

Conducted AI safety and LLM alignment work by developing and refining prompt/red-team scenarios for cybersecurity risks like prompt injection and jailbreaking. Evaluated text-based model outputs for accuracy and safety and authored rationales for ranking decisions and correction of hallucinations. Performed pairwise ranking, response auditing, and edge-case labeling to strengthen safety guardrails. • Developed technical cybersecurity/AI training content, including injection and jailbreaking cases • Ran independent red-teaming labs focused on prompt injection and jailbreaking techniques • Audited and improved model responses for accuracy and safety • Labeled edge cases and supported dataset quality via pairwise ranking

2025 - Present

AI Trainer (Translated Tool: Label Studio)

Label StudioTextRLHF

Executed LLM training workflows (SFT, RLHF, and DPO) to optimize model performance and alignment. Authored and applied training judgments and rationales to ensure outputs matched safety and alignment requirements. Conducted response ranking and auditing to improve the quality of the training datasets. • Performed SFT, RLHF, and DPO workflow executions • Produced logical rationales for ranking decisions and correcting hallucinations • Applied pairwise ranking and response auditing • Identified and labeled edge cases to improve safety guardrails

2024 - 2024

Education

Work History