Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
AWS SageMaker
Text
Computer Code Programming
Document
Computer Programming/Coding
Prompt + Response Writing (SFT)
Function Calling
Transcription
RLHF
Text Summarization
Question Answering
Text GenerationAs a Bachelor of Science graduate from McMaster University (2023), I bring a strong analytical foundation and extensive full-stack development experience to the field of AI training. I specialize in high-complexity code annotation, RLHF (Reinforcement Learning from Human Feedback), and SFT (Supervised Fine-Tuning) for modern programming languages and cloud-native frameworks. My hands-on experience with API integrations, database management, and deployment workflows allows me to identify subtle logical errors and "hallucinations" that generic models often miss, ensuring that model outputs align with professional production standards and best practices. Beyond technical coding, I have a proven track record in developing Standard Operating Procedures (SOPs) and technical documentation, which I leverage to create high-quality synthetic datasets and complex instruction-following tasks. I excel at Red Teaming and evaluating model responses for reasoning accuracy, safety, and structural integrity. My unique combination of technical leadership experience and scientific rigor enables me to bridge the gap between abstract prompts and precise, high-utility AI responses, making me an ideal collaborator for training models in specialized technical and operational domains.
Computer Code ProgrammingRLHF主要负责对 AI 生成的代码进行深度评估与对齐。针对 Python、JavaScript 以及 SQL 等语言，对模型输出的逻辑正确性、安全漏洞及执行效率进行打分。重点工作包括： 代码调试与纠错：识别模型在处理 Supabase 集成、PostgreSQL 复杂查询以及 Vercel 部署脚本时的“幻觉”现象。 偏好对齐 (RLHF)：在多个候选代码块中，根据可读性、模块化程度及是否符合现代开发最佳实践（Best Practices）进行排序，引导模型生成更优雅的生产级代码。
TextPrompt + Response Writing (SFT)参与高质量合成数据生成（Synthetic Data Generation）项目，专注于提升模型在处理企业级标准作业程序（SOP）时的推理能力。 SFT 数据撰写：模拟复杂业务场景（如企业合规性核查、运营流程自动化），撰写高质量的 Prompt-Response 对，确保模型输出符合特定的格式约束和逻辑严密性。 长文本摘要与重写：针对长篇技术协议和理赔文档进行精炼，标注核心关键实体（NER），并评估模型提取信息的准确度。 安全性评估：对模型生成的建议进行红队测试（Red Teaming），确保其在提供操作建议时符合法律法规与安全准则。
Bachelor of Science, Mathematics and Computer Science
cto
