
Outlier.ai
Remote AI Trainer
Feb 2025 — Present
CurrentRemote
- Contributed to large-scale model training pipelines by generating high-quality supervised data, evaluating model outputs, and refining decision boundaries
- Collaborated on LLM and CV model evaluation tasks, identifying failure modes and improving training data distributions
- Designed Python-based preprocessing and validation scripts to improve dataset consistency and model robustness
- Worked within containerized AI environments (Docker) and Git-controlled workflows to ensure reproducibility and scalable dataset generation





