Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
Label Studio
Argilla
Surge AI
Scale AI
Text
Document
Computer Code Programming
Red Teaming
Computer Programming/Coding
Prompt + Response Writing (SFT)
RLHF
Fine-tuning
Classification
Entity (NER) ClassificationExperienced AI Trainer and Data Labeler specializing in code review annotation, LLM output evaluation, red teaming, and adversarial prompt crafting. I have reviewed and annotated AI-generated and human-written code across Python, JavaScript, TypeScript, Solidity, and Bash — evaluating for correctness, security vulnerabilities, logic errors, and instruction-following accuracy. My red teaming work includes adversarial prompt design, jailbreak testing, threat modelling, and offensive security research applied to AI systems. I have evaluated LLM outputs across multiple providers for reasoning quality, factual accuracy, hallucination, and task completion — and created prompt-response pairs for agent fine-tuning. Additional experience in writing quality assessment, structured text annotation, and document labeling. Native English speaker with strong technical communication skills.
TextPrompt + Response Writing (SFT)Designed and wrote high-quality prompt-response pairs for supervised fine-tuning of language models. Covered technical domains including software engineering, cybersecurity, and Web3. Ensured outputs met instruction-following standards, factual accuracy requirements, and stylistic consistency guidelines.
TextRLHFEvaluated LLM-integrated agent responses for factual accuracy, reasoning quality, and task completion across multiple model providers. Curated prompt-response pairs by flagging off-task outputs, repetition loops, and instruction drift relevant to agent training and tuning workflows. Emphasized systematic assessment of LLM failure modes such as hallucinations and quality degradation. • Rated and assessed agent output correctness and reasoning quality • Created and curated prompt-response pairs for tuning • Identified off-task behavior, loops, and instruction drift • Worked with OpenAI, Claude, and Groq LLaMA 3 providers
Document
Evaluation/RatingEvaluated technical and general writing by labeling for clarity, coherence, instruction-following, and factual accuracy across multiple deliverables. Audited AI-generated writing for hallucinations, off-topic drift, tone mismatches, and formatting failures that would affect training or preference judgments. Applied a strong English proficiency baseline to support consistent writing quality ranking. • Labeled writing clarity, coherence, and instruction adherence • Checked for hallucinations, drift, and tone mismatches • Assessed formatting failures and output presentation • Evaluated both technical and general writing quality
TextRed TeamingCrafted adversarial prompts to probe AI system weaknesses, jailbreak resistance, and bias. Applied offensive security background to AI safety evaluation — including threat modelling, attack pattern analysis, and structured documentation of discovered vulnerabilities and failure modes.
Computer Code ProgrammingComputer Programming/CodingReviewed AI-generated and human-written code to identify correctness issues, security vulnerabilities, and logic/style problems across multiple programming and scripting languages. Produced structured remediation notes that can be reused as consistent labeling guidance. Focused on instruction-following accuracy, edge-case handling, and security anti-pattern detection (e.g., injection and replay-related vectors). • Annotated code quality at function and module level • Evaluated edge cases, security weaknesses, and logic errors • Documented findings with structured remediation notes • Covered Python, JavaScript/TypeScript, Solidity, and Bash
Bachelor of Science, Computer Science
IT Technical Support (National Service)
Security & Systems Analyst Intern
